OPM Data Separation Analysis¶

Christopher Boomhower₁, Stacey Fabricant₂, Alex Frye₁, David Mumford₂, Michael Smith₁, Lindsay Vitovsky₁

₁ Southern Methodist University, Dallas, TX, US ₂ Penn Mutual Life Insurance Co, Horsham PA </i></b>

Introduction¶

background text...

our intent is to: 1)..2)...3)........

Data Understanding¶

Data Source Background Text & citation links

Dataset Attribute Descriptions

Load the Data¶

To begin our analysis, we need to load the data from our 89 source .txt files. Data is separated into two separate groups of files; Separation and Non-Separation, thus data is loaded in two separate phases, then unioned together. Once data is loaded, Steps taken to remove non-US observations or those with no specified occupation, no specified salary, or no specified length of service level. Of a total 8,423,336 observations, we end with 8,232,375 after removal of these observations.

## Import libraries
import pickle
import os
import psutil
import glob
import pandas as pd
import numpy as np
from IPython.display import display
import matplotlib
import matplotlib.pyplot as plt
from matplotlib.colors import ListedColormap
import seaborn as sns
import requests
import json
import missingno as msno
import prettytable
import math
from sklearn.preprocessing import MinMaxScaler, StandardScaler, label_binarize
from sklearn.multiclass import OneVsRestClassifier
from sklearn.utils import class_weight
from sklearn.decomposition import PCA
from sklearn.pipeline import Pipeline
from sklearn.model_selection import StratifiedKFold
from sklearn.cross_validation import cross_val_score
from sklearn.metrics import roc_curve, auc
from sklearn.metrics import roc_auc_score
from scipy import interp
from sklearn.metrics import confusion_matrix
from sklearn.ensemble  import RandomForestClassifier
from sklearn.neighbors import KNeighborsClassifier
from sklearn.linear_model import LogisticRegression
from sklearn.cross_validation import ShuffleSplit
from sklearn.metrics import log_loss
from sklearn.metrics import roc_auc_score
from datetime import datetime
from dateutil.parser import parse
from itertools import cycle
from sklearn import metrics as mt
from sklearn.feature_selection import chi2
import itertools

#Need to make sure you install the rpy2 package via following command in the Putty genuse41 console:
#python3 /usr/bin/pip install --user rpy2
#NOTE: If the above pip install does not work, try the following instead:
#python3 /usr/local/es7/lib/python3.5/site-packages/pip install --user rpy2
%load_ext rpy2.ipython
from rpy2.robjects import pandas2ri


## Library Options

pd.options.mode.chained_assignment = None

pd.set_option('display.max_rows', 500)
pd.set_option('display.max_columns', 500)
pd.set_option('display.width', 1000)

/usr/local/es7/lib/python3.5/site-packages/sklearn/cross_validation.py:44: DeprecationWarning: This module was deprecated in version 0.18 in favor of the model_selection module into which all the refactored classes and functions are moved. Also note that the interface of the new CV iterators are different from that of this module. This module will be removed in 0.20.
  "This module will be removed in 0.20.", DeprecationWarning)

## Pre-defined Functions for use later
def pickleObject(objectname, filename, filepath = "PickleJar/"):
    fullpicklepath = "{0}{1}.pkl".format(filepath, filename)
    # Create a variable to pickle and open it in write mode
    picklefile = open(fullpicklepath, 'wb')
    pickle.dump(objectname, picklefile)
    picklefile.close()
    
def unpickleObject(filename, filepath = "PickleJar/"):
    fullunpicklepath = "{0}{1}.pkl".format(filepath, filename)
    # Create an variable to pickle and open it in write mode
    unpicklefile = open(fullunpicklepath, 'rb')
    unpickleObject = pickle.load(unpicklefile)
    unpicklefile.close()

    return unpickleObject
    
def clear_display():
    from IPython.display import clear_output
    
## Pre-defined variables for use later
dataOPMPath = "dataOPM"
dataEMPPath = "dataEMP"
PickleJarPath = "PickleJar"

%%time

## Load OPMSeparation Files

OPMDataFiles = glob.glob(os.path.join(dataOPMPath, "*.txt"))

for i in range(0,len(OPMDataFiles)):
    OPMDataFiles[i] = OPMDataFiles[i].replace("\\","/")

OPMDataList = []

for i,j in zip(OPMDataFiles,range(0,len(OPMDataFiles))):
    OPMDataList.append(pd.read_csv(i, dtype = 'str'))
    display(OPMDataList[j].head())

## Load the SEPDATA_FY2015 file into it's own object
indexes = [i for i,x in enumerate(OPMDataFiles) if x == 'dataOPM/SEPDATA_FY2015.txt']
OPMDataOrig = OPMDataList[indexes[0]]

CPU times: user 415 ms, sys: 39.1 ms, total: 454 ms
Wall time: 456 ms

%%time

#print(OPMDataFiles)

print(len(OPMDataOrig))

##### Merge / Modify Codes / Aggregate Attributes to be more descriptive per the metadata files

OPMDataMerged = OPMDataOrig.copy()

##AGYSUB - AGYTYP, AGY
indexes = [i for i,x in enumerate(OPMDataFiles) if x == 'dataOPM/DTagy.txt']
OPMDataMerged = OPMDataMerged.merge(OPMDataList[indexes[0]], on = 'AGYSUB', how = 'left')

##EFDate - quarter, month
indexes = [i for i,x in enumerate(OPMDataFiles) if x == 'dataOPM/DTefdate.txt']
OPMDataMerged = OPMDataMerged.merge(OPMDataList[indexes[0]], on = 'EFDATE', how = 'left')

##AGELVL - AGELVLT
indexes = [i for i,x in enumerate(OPMDataFiles) if x == 'dataOPM/DTagelvl.txt']
OPMDataMerged = OPMDataMerged.merge(OPMDataList[indexes[0]], on = 'AGELVL', how = 'left')

##LOSLVL - LOSLVLT
indexes = [i for i,x in enumerate(OPMDataFiles) if x == 'dataOPM/DTloslvl.txt']
OPMDataMerged = OPMDataMerged.merge(OPMDataList[indexes[0]], on = 'LOSLVL', how = 'left')

##LOC - LocTypeT, LocT
indexes = [i for i,x in enumerate(OPMDataFiles) if x == 'dataOPM/DTloc.txt']
OPMDataMerged = OPMDataMerged.merge(OPMDataList[indexes[0]], on = 'LOC', how = 'left')

##OCC - OCCTYPT, OCCFAM
indexes = [i for i,x in enumerate(OPMDataFiles) if x == 'dataOPM/DTocc.txt']
OPMDataMerged = OPMDataMerged.merge(OPMDataList[indexes[0]], on = 'OCC', how = 'left')

##PATCO - PATCOT
indexes = [i for i,x in enumerate(OPMDataFiles) if x == 'dataOPM/DTpatco.txt']
OPMDataMerged = OPMDataMerged.merge(OPMDataList[indexes[0]], on = 'PATCO', how = 'left')

##PPGRD - PayPlan, PPGroup, PPTYP
indexes = [i for i,x in enumerate(OPMDataFiles) if x == 'dataOPM/DTppgrd.txt']
OPMDataMerged = OPMDataMerged.merge(OPMDataList[indexes[0]], on = 'PPGRD', how = 'left')

##SALLVL - SALLVLT
indexes = [i for i,x in enumerate(OPMDataFiles) if x == 'dataOPM/DTsallvl.txt']
OPMDataMerged = OPMDataMerged.merge(OPMDataList[indexes[0]], on = 'SALLVL', how = 'left')

##TOA - TOATYP
indexes = [i for i,x in enumerate(OPMDataFiles) if x == 'dataOPM/DTtoa.txt']
OPMDataMerged = OPMDataMerged.merge(OPMDataList[indexes[0]], on = 'TOA', how = 'left')

##WORKSCH - WSTYPT
indexes = [i for i,x in enumerate(OPMDataFiles) if x == 'dataOPM/DTwrksch.txt']
OPMDataMerged = OPMDataMerged.merge(OPMDataList[indexes[0]], on = 'WORKSCH', how = 'left')


## Modify Data Types for numeric objects
OPMDataMerged["SALARY"] = OPMDataMerged["SALARY"].apply(pd.to_numeric)
OPMDataMerged["COUNT"]  = OPMDataMerged["COUNT"].apply(pd.to_numeric)
OPMDataMerged["LOS"]    = OPMDataMerged["LOS"].apply(pd.to_numeric)

print("Original SEP data size of: "+str(len(OPMDataMerged)))
print("Removing "+str(len(OPMDataMerged[OPMDataMerged["LOCTYP"] != "1"]))+" Non-US observations.")

## Remove Non-US Data
OPMDataMerged = OPMDataMerged[OPMDataMerged["LOCTYP"] == "1"]

print("Removing "+str(len(OPMDataMerged[OPMDataMerged["OCCTYP"] == "3"]))+" observations with no specified Occupation.")

   ## Remove Observations with no specified occupation
OPMDataMerged = OPMDataMerged[OPMDataMerged["OCCTYP"] != "3"]

print("Removing "+str(len(OPMDataMerged[OPMDataMerged["SALLVL"] == "Z"]))+" observations with no specified Salary.")

   ## Remove Observations with no specified salary
OPMDataMerged = OPMDataMerged[OPMDataMerged["SALLVL"] != "Z"]

print("Removing "+str(len(OPMDataMerged[OPMDataMerged["LOSLVL"] == "Z"]))+" observations with no specified Length of Service.")

   ## Remove Observations with no specified LOSLVL
OPMDataMerged = OPMDataMerged[OPMDataMerged["LOSLVL"] != "Z"]

print("Removing "+str(len(OPMDataMerged[OPMDataMerged["AGELVL"] == "A"]))+" observations of Age Level A")

## Remove Observations from Age Level A (less than 20 years old)
OPMDataMerged = OPMDataMerged[OPMDataMerged["AGELVL"] != "A"]

print("Removing "+str(len(OPMDataMerged[OPMDataMerged["AGELVL"] == "Z"]))+" observations with no specified Age Level.")

   ## Remove Observations with no specified Age Level
OPMDataMerged = OPMDataMerged[OPMDataMerged["AGELVL"] != "Z"]

    ## Fix differences in spaces on WORKSCHT Column
OPMDataMerged["WORKSCHT"] = np.where(OPMDataMerged["WORKSCHT"].str[0]=="F", 'Full-time Nonseasonal',
                                np.where(OPMDataMerged["WORKSCHT"].str[0]=="I", 'Intermittent Nonseasonal',
                                         np.where(OPMDataMerged["WORKSCHT"].str[0]=="P", 'Part-time Nonseasonal',
                                                  np.where(OPMDataMerged["WORKSCHT"].str[0]=="G", 'Full-time Seasonal',
                                                        np.where(OPMDataMerged["WORKSCHT"].str[0]=="J", 'Intermittent Seasonal',
                                                                np.where(OPMDataMerged["WORKSCHT"].str[0]=="Q", 'Part-time Seasonal',
                                                                        np.where(OPMDataMerged["WORKSCHT"].str[0]=="T", 'Part-time Job Sharer Seasonal',
                                                                                np.where(OPMDataMerged["WORKSCHT"].str[0]=="S", 'Part-time Job Sharer Nonseasonal',
                                                                                        np.where(OPMDataMerged["WORKSCHT"].str[0]=="B", 'Full-time Nonseasonal Baylor Plan',
                                                                                                'NO WORK SCHEDULE REPORTED' ### ELSE case represents Night
                                                                                                 )
                                                                                         )
                                                                                 )
                                                                         )
                                                                 )
                                                          )
                                                 )
                                        )
                               )    

display(OPMDataMerged.head())
print("New SEP data size of: "+str(len(OPMDataMerged)))
display(OPMDataMerged.describe().transpose())
#del OPMDataList,OPMDataFiles

226357
Original SEP data size of: 226357
Removing 8021 Non-US observations.
Removing 55 observations with no specified Occupation.
Removing 1426 observations with no specified Salary.
Removing 3 observations with no specified Length of Service.
Removing 2570 observations of Age Level A
Removing 0 observations with no specified Age Level.

New SEP data size of: 214282

CPU times: user 14.5 s, sys: 203 ms, total: 14.7 s
Wall time: 14.7 s

%%time

if os.path.isfile(PickleJarPath+"/EMPDataOrig4Q.pkl"):
    print("Found the File! Loading Pickle Now!")
    EMPDataOrig4Q = unpickleObject("EMPDataOrig4Q")
else:
    ## Load EMPData Files

    indexes = []
    EMPDataFiles = []
    EMPDataList = []
    EMPDataOrig = []

    for i,qtr in enumerate(["Q1", "Q2", "Q3", "Q4"]): 
        EMPDataFiles.append(glob.glob(os.path.join(dataEMPPath, qtr + "/*.txt")))

        for j in range(0,len(EMPDataFiles[i])):
            EMPDataFiles[i][j] = EMPDataFiles[i][j].replace("\\","/")

        EMPDataList.append([])

        for j,file in enumerate(EMPDataFiles[i]):
            EMPDataList[i].append(pd.read_csv(file, dtype = 'str'))
            if i == 0:
                display(EMPDataList[i][j].head())

        ## Load the FactData files into it's own object
        indexes.append([])
            ##[qtr][fileindex from EMPDataList]
        indexes[i]=[j for j,x in enumerate(EMPDataFiles[i]) if dataEMPPath + '/' + qtr + '/FACTDATA' in x]   

        EMPDataOrig.append([])

        EMPDataOrig[i] = pd.concat([EMPDataList[i][indexes[i][j]] for j in range(0,len(indexes[i]))]) 
        EMPDataOrig[i]["QTR"] = str(i+1)

            ## modify data type for numerics
        EMPDataOrig[i]["SALARY"] = EMPDataOrig[i]["SALARY"].str.replace(',', '').str.replace('$', '').str.replace(' ', '').apply(pd.to_numeric)
      
        ## Load Metadata
        ##AGYSUB - AGYTYP, AGY
        ind2 = [i for i,x in enumerate(EMPDataFiles[i]) if x == dataEMPPath + '/' + qtr + '/DTagy.txt']
        EMPDataOrig[i] = EMPDataOrig[i].merge(EMPDataList[i][ind2[0]], on = 'AGYSUB', how = 'left')

        ##AGELVL - AGELVLT
        ind2 = [i for i,x in enumerate(EMPDataFiles[i]) if x == dataEMPPath + '/' + qtr + '/DTagelvl.txt']
        EMPDataOrig[i] = EMPDataOrig[i].merge(EMPDataList[i][ind2[0]], on = 'AGELVL', how = 'left')

        #LOSLVL - LOSLVLT
        ind2 = [i for i,x in enumerate(EMPDataFiles[i]) if x == dataEMPPath + '/' + qtr + '/DTloslvl.txt']
        EMPDataOrig[i] = EMPDataOrig[i].merge(EMPDataList[i][ind2[0]], on = 'LOSLVL', how = 'left')
        EMPDataOrig[i]["LOS"] = EMPDataOrig[i]["LOS"].apply(pd.to_numeric)
        
        ##LOC - LocTypeT, LocT
        ind2 = [i for i,x in enumerate(EMPDataFiles[i]) if x == dataEMPPath + '/' + qtr + '/DTloc.txt']
        EMPDataOrig[i] = EMPDataOrig[i].merge(EMPDataList[i][ind2[0]], on = 'LOC', how = 'left')
 
        ##OCC - OCCTYPT, OCCFAM
        ind2 = [i for i,x in enumerate(EMPDataFiles[i]) if x == dataEMPPath + '/' + qtr + '/DTocc.txt']
        EMPDataOrig[i] = EMPDataOrig[i].merge(EMPDataList[i][ind2[0]], on = 'OCC', how = 'left')

        ##PATCO - PATCOT
        ind2 = [i for i,x in enumerate(EMPDataFiles[i]) if x == dataEMPPath + '/' + qtr + '/DTpatco.txt']
        EMPDataOrig[i] = EMPDataOrig[i].merge(EMPDataList[i][ind2[0]], on = 'PATCO', how = 'left')

        ##PPGRD - PayPlan, PPGroup, PPTYP
        ind2 = [i for i,x in enumerate(EMPDataFiles[i]) if x == dataEMPPath + '/' + qtr + '/DTppgrd.txt']
        EMPDataOrig[i] = EMPDataOrig[i].merge(EMPDataList[i][ind2[0]], on = 'PPGRD', how = 'left')

        ##SALLVL - SALLVLT
        ind2 = [i for i,x in enumerate(EMPDataFiles[i]) if x == dataEMPPath + '/' + qtr + '/DTsallvl.txt']
        EMPDataOrig[i] = EMPDataOrig[i].merge(EMPDataList[i][ind2[0]], on = 'SALLVL', how = 'left')

        ##TOA - TOATYP
        ind2 = [i for i,x in enumerate(EMPDataFiles[i]) if x == dataEMPPath + '/' + qtr + '/DTtoa.txt']
        EMPDataOrig[i] = EMPDataOrig[i].merge(EMPDataList[i][ind2[0]], on = 'TOA', how = 'left')

        ##WORKSCH - WSTYPT
        ind2 = [i for i,x in enumerate(EMPDataFiles[i]) if x == dataEMPPath + '/' + qtr + '/DTwrksch.txt']
        EMPDataOrig[i] = EMPDataOrig[i].merge(EMPDataList[i][ind2[0]], on = 'WORKSCH', how = 'left')

        display(EMPDataOrig[i].head())

    EMPDataOrig4Q = pd.concat([EMPDataOrig[j] for j in range(0,len(EMPDataOrig))])
    print("Original EMP data size of: "+str(len(EMPDataOrig4Q)))
    print("Removing "+str(len(EMPDataOrig4Q[EMPDataOrig4Q["LOCTYP"] != "1"]))+" Non-US observations.")
    
       ## Remove Non-US Data
    EMPDataOrig4Q = EMPDataOrig4Q[EMPDataOrig4Q["LOCTYP"] == "1"]

    print("Removing "+str(len(EMPDataOrig4Q[EMPDataOrig4Q["OCCTYP"] == "3"]))+" observations with no specified Occupation.")

       ## Remove Observations with no specified occupation
    EMPDataOrig4Q = EMPDataOrig4Q[EMPDataOrig4Q["OCCTYP"] != "3"]

    print("Removing "+str(len(EMPDataOrig4Q[EMPDataOrig4Q["SALLVL"] == "Z"]))+" observations with no specified Salary.")

       ## Remove Observations with no specified salary
    EMPDataOrig4Q = EMPDataOrig4Q[EMPDataOrig4Q["SALLVL"] != "Z"]

    print("Removing "+str(len(EMPDataOrig4Q[EMPDataOrig4Q["LOSLVL"] == "Z"]))+" observations with no specified Length of Service.")

       ## Remove Observations with no specified LOSLVL
    EMPDataOrig4Q = EMPDataOrig4Q[EMPDataOrig4Q["LOSLVL"] != "Z"]

    print("Removing "+str(len(EMPDataOrig4Q[EMPDataOrig4Q["AGELVL"] == "A"]))+" observations of Age Level A.")

        ## Remove Observations from Age Level A (less than 20 years old)
    EMPDataOrig4Q = EMPDataOrig4Q[EMPDataOrig4Q["AGELVL"] != "A"]

    print("Removing "+str(len(EMPDataOrig4Q[EMPDataOrig4Q["AGELVL"] == "Z"]))+" observations with no specified Age Level.")

        ## Remove Observations with no specified Age Level
    EMPDataOrig4Q = EMPDataOrig4Q[EMPDataOrig4Q["AGELVL"] != "Z"]

        ## Fix differences in spaces on WORKSCHT Column
    EMPDataOrig4Q["WORKSCHT"] = np.where(EMPDataOrig4Q["WORKSCHT"].str[0]=="F", 'Full-time Nonseasonal',
                                    np.where(EMPDataOrig4Q["WORKSCHT"].str[0]=="I", 'Intermittent Nonseasonal',
                                             np.where(EMPDataOrig4Q["WORKSCHT"].str[0]=="P", 'Part-time Nonseasonal',
                                                      np.where(EMPDataOrig4Q["WORKSCHT"].str[0]=="G", 'Full-time Seasonal',
                                                            np.where(EMPDataOrig4Q["WORKSCHT"].str[0]=="J", 'Intermittent Seasonal',
                                                                    np.where(EMPDataOrig4Q["WORKSCHT"].str[0]=="Q", 'Part-time Seasonal',
                                                                            np.where(EMPDataOrig4Q["WORKSCHT"].str[0]=="T", 'Part-time Job Sharer Seasonal',
                                                                                    np.where(EMPDataOrig4Q["WORKSCHT"].str[0]=="S", 'Part-time Job Sharer Nonseasonal',
                                                                                            np.where(EMPDataOrig4Q["WORKSCHT"].str[0]=="B", 'Full-time Nonseasonal Baylor Plan',
                                                                                                    'NO WORK SCHEDULE REPORTED' ### ELSE case represents Night
                                                                                                     )
                                                                                             )
                                                                                     )
                                                                             )
                                                                     )
                                                              )
                                                     )
                                            )
                                   )    

    pickleObject(EMPDataOrig4Q, "EMPDataOrig4Q")

print("New EMP data size of: "+str(len(EMPDataOrig4Q)))

Found the File! Loading Pickle Now!
New EMP data size of: 8008911
CPU times: user 9.06 s, sys: 1.76 s, total: 10.8 s
Wall time: 10.8 s

display(EMPDataOrig4Q.describe().transpose())

%matplotlib inline

#sns.boxplot(y = "SALARY", data = EMPDataOrig4Q)

With both our separation and non-separation data loaded, we calculate three new attributes through aggregation or calculation amongst various attributes.

1) SEP Count by Date & Occupation – total number of separations (of any type) for a given Date and Occupation;

2) SEP Count by Date & Location – total number of separations (of any type) for a given Date and Location;

3) Industry Average Salary – Average salary amongst non-separated employees, grouped by quarter, occupation, pay grade, and work schedule;

We proceed, by concatenating our Separation and Non-Separation observations, and merge these newly calculated attributes to the concatenated dataset.

%%time
%matplotlib inline

##Aggregate Number of Total Separations in current month for given Occ
AggSEPCount_EFDATE_OCC= pd.DataFrame({'SEPCount_EFDATE_OCC' : OPMDataMerged.groupby(["EFDATE", "OCC"]).size()}).reset_index()
display(AggSEPCount_EFDATE_OCC.head())


##Aggregate Number of Total Separations in current month for given LOC
AggSEPCount_EFDATE_LOC = pd.DataFrame({'SEPCount_EFDATE_LOC' : OPMDataMerged.groupby(["EFDATE", "LOC"]).size()}).reset_index()
display(AggSEPCount_EFDATE_LOC.head())

##Average Quarterly EMP Salary by occ 
AggIndAvgSalary = pd.DataFrame({'count' : EMPDataOrig4Q.groupby(["QTR", "OCC", "PPGRD", "WORKSCHT"]).size()}).reset_index()
AggIndAvgSalary2 = pd.DataFrame({'IndSalarySum' : EMPDataOrig4Q.groupby(["QTR", "OCC", "PPGRD", "WORKSCHT"])["SALARY"].sum()}).reset_index()
AggIndAvgSalary = AggIndAvgSalary.merge(AggIndAvgSalary2,on=["QTR", "OCC", "PPGRD", "WORKSCHT"])
AggIndAvgSalary["IndAvgSalary"] = AggIndAvgSalary["IndSalarySum"]/AggIndAvgSalary["count"]
del AggIndAvgSalary["count"]
del AggIndAvgSalary["IndSalarySum"]
display(AggIndAvgSalary.head())

CPU times: user 3.02 s, sys: 465 ms, total: 3.48 s
Wall time: 3.48 s

#Merge Two Datasets
### NS SEP code means NonSeparation
###add hardcoded null value columns where applicable
EMPDataOrig4Q["SEP"] = "NS"
EMPDataOrig4Q["GENDER"] = np.nan
EMPDataOrig4Q["COUNT"] = np.nan

OPMDataMerged["DATECODE"] = OPMDataMerged["EFDATE"]

OPMColList = ["AGYSUB", "SEP", "DATECODE",   "AGELVL", "GENDER", "GSEGRD", "LOSLVL", "LOC", "OCC", "PATCO", "PPGRD", "SALLVL", "TOA", "WORKSCH", "COUNT", "SALARY", "LOS", "AGYTYP", "AGYTYPT", "AGY", "AGYT", "AGYSUBT", "QTR", "AGELVLT", "LOSLVLT", "LOCTYP", "LOCTYPT", "LOCT", "OCCTYP", "OCCTYPT", "OCCFAM", "OCCFAMT", "OCCT", "PATCOT", "PPTYP", "PPTYPT", "PPGROUP", "PPGROUPT", "PAYPLAN", "PAYPLANT", "SALLVLT", "TOATYP", "TOATYPT", "TOAT", "WSTYP", "WSTYPT", "WORKSCHT"]
EMPColList = ["AGYSUB", "SEP", "DATECODE", "AGELVL", "GENDER", "GSEGRD", "LOSLVL", "LOC", "OCC", "PATCO", "PPGRD", "SALLVL", "TOA", "WORKSCH", "COUNT", "SALARY", "LOS", "AGYTYP", "AGYTYPT", "AGY", "AGYT", "AGYSUBT", "QTR", "AGELVLT", "LOSLVLT", "LOCTYP", "LOCTYPT", "LOCT", "OCCTYP", "OCCTYPT", "OCCFAM", "OCCFAMT", "OCCT", "PATCOT", "PPTYP", "PPTYPT", "PPGROUP", "PPGROUPT", "PAYPLAN", "PAYPLANT", "SALLVLT", "TOATYP", "TOATYPT", "TOAT", "WSTYP", "WSTYPT", "WORKSCHT"]

OPMDataMerged = pd.concat([OPMDataMerged[OPMColList], EMPDataOrig4Q[EMPColList]], ignore_index=True)
print("Total concatenated data size for SEP and non-SEP: "+str(len(OPMDataMerged)))

OPMDataMerged = OPMDataMerged.merge(AggSEPCount_EFDATE_OCC, left_on = ['DATECODE','OCC'], right_on = ['EFDATE','OCC'], how = 'left')
OPMDataMerged = OPMDataMerged.merge(AggSEPCount_EFDATE_LOC, left_on = ['DATECODE','LOC'], right_on = ['EFDATE','LOC'], how = 'left')
OPMDataMerged = OPMDataMerged.merge(AggIndAvgSalary, on = ['QTR','OCC', 'PPGRD', 'WORKSCHT'], how = 'left')
OPMDataMerged["SalaryOverUnderIndAvg"] = OPMDataMerged["SALARY"] - OPMDataMerged["IndAvgSalary"]

del OPMDataMerged["EFDATE_x"]
del OPMDataMerged["EFDATE_y"]

display(OPMDataMerged.head())
display(OPMDataMerged.tail())

Total concatenated data size for SEP and non-SEP: 8223193

print(len(OPMDataMerged[OPMDataMerged["SEPCount_EFDATE_OCC"].isnull()]))

display(OPMDataMerged[OPMDataMerged["SEPCount_EFDATE_OCC"].isnull()][["SEP","DATECODE", "OCC"]].drop_duplicates())

50993

These 50993 Non-Separation observations do not have coverage within the Separation Dataset, thus, we will remove these observations as out of scope demographic in our analysis. Any attempt in predicting these values will not have enough data to support a significant response.

OPMDataMerged = OPMDataMerged[OPMDataMerged["SEPCount_EFDATE_OCC"].notnull()]

print(len(OPMDataMerged[OPMDataMerged["SEPCount_EFDATE_OCC"].isnull()]))

print(len(OPMDataMerged))

0
8172200

print(len(OPMDataMerged[OPMDataMerged["SEPCount_EFDATE_LOC"].isnull()]))

display(OPMDataMerged[OPMDataMerged["SEPCount_EFDATE_LOC"].isnull()][["SEP","DATECODE","LOC"]].drop_duplicates())

0

print(len(OPMDataMerged[OPMDataMerged["IndAvgSalary"].isnull()]))

display(OPMDataMerged[OPMDataMerged["IndAvgSalary"].isnull()][["QTR", "SEP","OCCT", "PPGRD", "WORKSCHT"]].drop_duplicates())

1293

These 1293 separation observations do not have coverage within the EMP Dataset, thus, we will remove these observations as out of scope demographic in our analysis. Any attempt in predicting these values will not have enough data to support a significant response.

OPMDataMerged = OPMDataMerged[OPMDataMerged["IndAvgSalary"].notnull()]

print(len(OPMDataMerged[OPMDataMerged["IndAvgSalary"].isnull()]))

print(len(OPMDataMerged))

0
8170907

Placeholder Chunks for Data Quality check of salary against GS Grade Level Ranges¶

# Placeholder Chunks for Data Quality check of salary against GS Grade Level Ranges

We are iterested to see how federal pension plans may impact attrition in this dataset. An interesting attribute to complement Length of service, is Years to Retirement. Utilizing a FERS retirement eligibility baseline of 57 years of age for all observations, and the lower limitation of age level ranges we compute a numeric value for length of retirement.

#Add Column YearsToRetirement

"""
    AGELVL,AGELVLT
    A,Less than 20
    B,20-24
    C,25-29
    D,30-34
    E,35-39
    F,40-44
    G,45-49
    H,50-54
    I,55-59
    J,60-64
    K,65 or more
    Z,Unspecified
"""
OPMDataMerged["LowerLimitAge"] = np.where(OPMDataMerged["AGELVL"]=="B", 20,
                                                np.where(OPMDataMerged["AGELVL"]=="C", 25,
                                                         np.where(OPMDataMerged["AGELVL"]=="D", 30,
                                                                  np.where(OPMDataMerged["AGELVL"]=="E", 35,
                                                                           np.where(OPMDataMerged["AGELVL"]=="F", 40,
                                                                                    np.where(OPMDataMerged["AGELVL"]=="G", 45,
                                                                                             np.where(OPMDataMerged["AGELVL"]=="H", 50,
                                                                                                      np.where(OPMDataMerged["AGELVL"]=="I", 55,
                                                                                                               np.where(OPMDataMerged["AGELVL"]=="J", 60,
                                                                                                                        np.where(OPMDataMerged["AGELVL"]=="K", 65,
                                                                                                                                 np.nan
                                                                                                                                )
                                                                                                                        )
                                                                                                               )
                                                                                                      )
                                                                                            )
                                                                                   )
                                                                          )
                                                                 )
                                                        )
                                               )  

retAge = 57

OPMDataMerged["YearsToRetirement"] = np.where(OPMDataMerged["AGELVL"]=="B", retAge-20,
                                                np.where(OPMDataMerged["AGELVL"]=="C", retAge-25,
                                                         np.where(OPMDataMerged["AGELVL"]=="D", retAge-30,
                                                                  np.where(OPMDataMerged["AGELVL"]=="E", retAge-35,
                                                                           np.where(OPMDataMerged["AGELVL"]=="F", retAge-40,
                                                                                    np.where(OPMDataMerged["AGELVL"]=="G", retAge-45,
                                                                                             np.where(OPMDataMerged["AGELVL"]=="H", retAge-50,
                                                                                                      np.where(OPMDataMerged["AGELVL"]=="I", retAge-55,
                                                                                                               np.where(OPMDataMerged["AGELVL"]=="J", retAge-60,
                                                                                                                        np.where(OPMDataMerged["AGELVL"]=="K", retAge-65,
                                                                                                                                 np.nan
                                                                                                                                )
                                                                                                                        )
                                                                                                               )
                                                                                                      )
                                                                                            )
                                                                                   )
                                                                          )
                                                                 )
                                                        )
                                               )  

print("Null Values for LowerLimitAge: " + str(len(OPMDataMerged[OPMDataMerged["LowerLimitAge"].isnull()])))
print("Null Values for YearsToRetirement: " + str(len(OPMDataMerged[OPMDataMerged["YearsToRetirement"].isnull()])))

display(OPMDataMerged.head())
display(OPMDataMerged.tail())

Null Values for LowerLimitAge: 0
Null Values for YearsToRetirement: 0

Pull Bureau of Labor Statistics data¶

In addition to the OPM data, we merge 10 attributes from the Bureau of Labor Statistics (BLS). Data is sourced from Federal Government industry codes across all regions. Although assumed to be highly correlated, we source both Level (Total number) and Rate (Percentage of Level to total employment and / or job openings) for the following statistics: 1) Job Openings, 2) Layoffs, 3) Quits, 4) Total Separations, and 5) Other Separations. While Rate paints an aggregated, holistic picture for job market trends, Level provides a raw count for total separations alone. Both these statistics were captured by a monthly aggregate and merged to the OPM data by their respective months.

%%time

def bls(series, start, end):
    headers = {'Content-type': 'application/json'}
    sID   = []
    
    for i in range(0,len(series)):
        sID.append(series[i][0])
    
    data = json.dumps({"seriesid": sID,
                       "startyear":start,
                       "endyear":end,
                       "catalog":False,
                       "calculations":False,
                       "annualaverage":False,
                       "registrationkey":"7a89c8d7979349fba8914b8be16a1646"})
    
    p = requests.post('https://api.bls.gov/publicAPI/v2/timeseries/data/', data=data, headers=headers)
    json_data = json.loads(p.text)
    bls = []
    for series in json_data['Results']['series']:
        #x=prettytable.PrettyTable(["series id","year","period","value","footnotes"])
        result = pd.DataFrame(columns=["series id","year","period","value","footnotes"])
        seriesId = series['seriesID']
        for item in series['data']:
            year = item['year']
            period = item['period']
            value = item['value']
            footnotes=""
            for footnote in item['footnotes']:
                if footnote:
                    footnotes = footnotes + footnote['text'] + ','
            if 'M01' <= period <= 'M12':
                #x.add_row([seriesId,year,period,value,footnotes[0:-1]])
                y = pd.DataFrame({"series id" : seriesId,
                                  "year" : year,
                                  "period" : period,
                                  "value" : value,
                                  "footnotes" : footnotes}, index = [0])
                result = result.append(y, ignore_index = True)
        bls.append(result)
    return(bls)

CPU times: user 2 µs, sys: 0 ns, total: 2 µs
Wall time: 7.15 µs

%%time

seriesList = [
              ['JTU91000000JOL','BLS_FEDERAL_JobOpenings_Level'],
              ['JTU91000000LDL','BLS_FEDERAL_Layoffs_Level'],
              ['JTU91000000OSL','BLS_FEDERAL_OtherSep_Level'],
              ['JTU91000000QUL','BLS_FEDERAL_Quits_Level'],
              ['JTU91000000TSL','BLS_FEDERAL_TotalSep_Level'],
              ['JTU91000000JOR','BLS_FEDERAL_JobOpenings_Rate'],
              ['JTU91000000LDR','BLS_FEDERAL_Layoffs_Rate'],
              ['JTU91000000OSR','BLS_FEDERAL_OtherSep_Rate'],
              ['JTU91000000QUR','BLS_FEDERAL_Quits_Rate'],
              ['JTU91000000TSR','BLS_FEDERAL_TotalSep_Rate']
             ]

# Pull job openings and labor turnover data
JTL = bls(seriesList, "2014", "2015")

seriesList = pd.DataFrame(seriesList, columns = ["series id","sName"])

##We need to replace these with actual Descriptor Column Names

for i in range(0,len(seriesList)):
    
    JTL[i] = JTL[i].merge(seriesList, on = "series id", how = 'inner')

    if len(JTL[i]) >0:
        name = JTL[i]["sName"].drop_duplicates().values[0]
    else:
        name = str(i)

    JTL[i][name] = JTL[i]["value"].apply(pd.to_numeric)
    JTL[i]["DATECODE"] = JTL[i]["year"] + JTL[i]["period"].str[-2:]
    del JTL[i]["value"]
    del JTL[i]["year"]
    del JTL[i]["period"]
    del JTL[i]["series id"]
    del JTL[i]["footnotes"]
    del JTL[i]["sName"]
    
    
    OPMDataMerged = OPMDataMerged.merge(JTL[i], on = "DATECODE", how = 'left')
    display(JTL[i].head())

CPU times: user 37.4 s, sys: 10.1 s, total: 47.4 s
Wall time: 47.9 s

display(OPMDataMerged.head())
display(OPMDataMerged.tail())

display(pd.DataFrame({'StratCount' : OPMDataMerged.groupby(["SEP"]).size()}).reset_index())

Preliminary EDA¶

In terms of data exploration, we first investigate numeric type attributes. Relationships, distributions, and correlation values are reviewed.

A new binary separation attribute is created to indicate whether non-sep or sep for EDA correlation purposes

#%%time
#
#
#cols = list(SampledOPMData.select_dtypes(include=['float64', 'int64']))
#cols.remove('COUNT')
#cols.remove('BLS_FEDERAL_OtherSep_Rate')
#cols.remove('BLS_FEDERAL_Quits_Rate')
#cols.remove('BLS_FEDERAL_TotalSep_Level')
#cols.remove('BLS_FEDERAL_JobOpenings_Rate')
#cols.remove('BLS_FEDERAL_OtherSep_Level')
#cols.remove('BLS_FEDERAL_Quits_Level')
#cols.remove('BLS_FEDERAL_JobOpenings_Level')
#cols.remove('BLS_FEDERAL_Layoffs_Rate')
#cols.remove('BLS_FEDERAL_Layoffs_Level')
#cols.remove('BLS_FEDERAL_TotalSep_Rate')
#cols.append('SEP')
#display(cols)
#
#plotNumeric = SampledOPMData[cols]
#
## Create binary separation attribute for EDA correlation review
##plotNumeric["SEP_bin"] = plotNumeric.SEP.replace("NS", 1)
##plotNumeric.loc[plotNumeric['SEP_bin'] != 1, 'SEP_bin'] = 0
##plotNumeric.SEP_bin = plotNumeric.SEP_bin.apply(pd.to_numeric)
#AttSplit = pd.get_dummies(plotNumeric['SEP'],prefix='SEP')
#display(AttSplit.head())
#plotNumeric = pd.concat((plotNumeric,AttSplit),axis=1) # add back into the dataframe
#
#display(plotNumeric.head())
#print("plotNumeric has {0} Records".format(len(plotNumeric)))
##print(plotNumeric.SEP_bin.dtype)

%%time


cols = list(OPMDataMerged.select_dtypes(include=['float64', 'int64']))
cols.remove('COUNT')
cols.remove('BLS_FEDERAL_OtherSep_Rate')
cols.remove('BLS_FEDERAL_Quits_Rate')
cols.remove('BLS_FEDERAL_TotalSep_Level')
cols.remove('BLS_FEDERAL_JobOpenings_Rate')
cols.remove('BLS_FEDERAL_OtherSep_Level')
cols.remove('BLS_FEDERAL_Quits_Level')
cols.remove('BLS_FEDERAL_JobOpenings_Level')
cols.remove('BLS_FEDERAL_Layoffs_Rate')
cols.remove('BLS_FEDERAL_Layoffs_Level')
cols.remove('BLS_FEDERAL_TotalSep_Rate')
cols.append('SEP')
display(cols)

plotNumeric = OPMDataMerged[cols]

# Create binary separation attribute for EDA correlation review
#plotNumeric["SEP_bin"] = plotNumeric.SEP.replace("NS", 1)
#plotNumeric.loc[plotNumeric['SEP_bin'] != 1, 'SEP_bin'] = 0
#plotNumeric.SEP_bin = plotNumeric.SEP_bin.apply(pd.to_numeric)
AttSplit = pd.get_dummies(plotNumeric['SEP'],prefix='SEP')
display(AttSplit.head())
plotNumeric = pd.concat((plotNumeric,AttSplit),axis=1) # add back into the dataframe

display(plotNumeric.head())
print("plotNumeric has {0} Records".format(len(plotNumeric)))
#print(plotNumeric.SEP_bin.dtype)

['SALARY',
 'LOS',
 'SEPCount_EFDATE_OCC',
 'SEPCount_EFDATE_LOC',
 'IndAvgSalary',
 'SalaryOverUnderIndAvg',
 'LowerLimitAge',
 'YearsToRetirement',
 'SEP']

plotNumeric has 8170907 Records
CPU times: user 1 s, sys: 510 ms, total: 1.51 s
Wall time: 1.51 s

#%%time
#
#sns.set(font_scale=1)
#sns.pairplot(plotNumeric.drop(["SEP_NS", 
#                               "SEP_SA", 
#                               "SEP_SB", 
#                               "SEP_SC", 
#                               "SEP_SD", 
#                               "SEP_SE", 
#                               "SEP_SF", 
#                               "SEP_SG", 
#                               "SEP_SH", 
#                               "SEP_SI", 
#                               "SEP_SJ", 
#                               "SEP_SK", 
#                               "SEP_SL"
#                              ], axis=1), hue = 'SEP', palette="hls", plot_kws={"s": 50})

%%time

# Function modified from https://stackoverflow.com/questions/29530355/plotting-multiple-histograms-in-grid
sns.set()

def draw_histograms(df, variables, n_rows, n_cols):
    fig=plt.figure(figsize=(20,20))
    for i, var_name in enumerate(variables):
        ax=fig.add_subplot(n_rows,n_cols,i+1)
        df[var_name].hist(bins=20,ax=ax, color='#58D68D')
        ax.set_title(var_name+" Distribution")
    fig.tight_layout()  # Improves appearance a bit.
    plt.show()

draw_histograms(plotNumeric.drop(['SEP',
                                  "SEP_NS", 
                               "SEP_SA", 
                               "SEP_SB", 
                               "SEP_SC", 
                               "SEP_SD", 
                               "SEP_SE", 
                               "SEP_SF", 
                               "SEP_SG", 
                               "SEP_SH", 
                               "SEP_SI", 
                               "SEP_SJ", 
                               "SEP_SK", 
                               "SEP_SL"
                              ], axis=1),
                plotNumeric.drop(['SEP',
                                  "SEP_NS", 
                               "SEP_SA", 
                               "SEP_SB", 
                               "SEP_SC", 
                               "SEP_SD", 
                               "SEP_SE", 
                               "SEP_SF", 
                               "SEP_SG", 
                               "SEP_SH", 
                               "SEP_SI", 
                               "SEP_SJ", 
                               "SEP_SK", 
                               "SEP_SL"
                              ], axis=1).columns, 6, 3)

/usr/local/es7/lib/python3.5/site-packages/matplotlib/font_manager.py:1297: UserWarning: findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans
  (prop.get_family(), self.defaultFamily[fontext]))

CPU times: user 4.68 s, sys: 6.2 s, total: 10.9 s
Wall time: 4.34 s

%%time
# Inspired by http://seaborn.pydata.org/examples/many_pairwise_correlations.html

#plt.matshow(plotNumeric.corr())

sns.set(style='white')
corr = plotNumeric.drop(['SEP'], axis=1).corr()

# Generate a mask for the upper triangle
mask = np.zeros_like(corr, dtype=np.bool)
mask[np.triu_indices_from(mask, k=1)] = True

# Set up the matplotlib figure
f, ax = plt.subplots(figsize=(20, 20))

# Generate a custom diverging colormap
cmap = sns.diverging_palette(250, 10, as_cmap=True)

# Draw the heatmap with the mask and correct aspect ratio
sns.set(font_scale=0.95)
heatCorr = sns.heatmap(corr, mask=mask, cmap=cmap, vmax=1, vmin=-1,
                       square=True, annot=True, linewidths=1,
                       cbar_kws={"shrink": .5}, ax=ax, fmt='.1g')
#heatCorr.
ax.tick_params(labelsize=15)
cax = plt.gcf().axes[-1]
cax.tick_params(labelsize=15)

sns.plt.show()
#sns.heatmap(corr, annot=True, linewidths=0.01, cmap=cmap, ax=ax)

/usr/local/es7/lib/python3.5/site-packages/matplotlib/font_manager.py:1297: UserWarning: findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans
  (prop.get_family(), self.defaultFamily[fontext]))

CPU times: user 8.74 s, sys: 1.27 s, total: 10 s
Wall time: 8.97 s

Based on the distribution of attributes identified above, we have decided to take the log transform of several attributes.

Salary
LOS (augmented by a value of .00001 to adjust for the undefined result of log(0)
SEPCount_EFDATE_OCC
SEPCount_EFDATE_LOC

%%time

# Log Transform Columns Added
OPMDataMerged["SALARYLog"] = OPMDataMerged["SALARY"].apply(np.log)
#OPMDataMerged["LOSLog"] = (OPMDataMerged["LOS"] + .00001).apply(np.log)
OPMDataMerged["LOSSqrt"] = (OPMDataMerged["LOS"]).apply(np.sqrt)
OPMDataMerged["SEPCount_EFDATE_OCCLog"] = OPMDataMerged["SEPCount_EFDATE_OCC"].apply(np.log)
OPMDataMerged["SEPCount_EFDATE_LOCLog"] = OPMDataMerged["SEPCount_EFDATE_LOC"].apply(np.log)
OPMDataMerged["IndAvgSalaryLog"] = OPMDataMerged["IndAvgSalary"].apply(np.log)

CPU times: user 1.05 s, sys: 116 ms, total: 1.17 s
Wall time: 1.17 s

We next review categorical data to improve our understanding of factor levels.

#%%time
#            "LOCTYPT",
#            "OCCTYP",
#            "OCCTYPT",
#            "PPTYP",
#            "PPTYPT",
#            "AGYTYP",
#            "OCCFAM",
#            "PPGROUP",
#            "PAYPLAN",
#            "TOATYP",
#            "WSTYP",
#            "AGYSUBT",
#            "AGELVL",
#            "LOSLVL",
#            "LOC",
#            "OCC",
#            "PATCO",
#            "SALLVL",
#            "TOA",
#            "WORKSCH"]
#
#for i in dropCols:
#    cols.remove(i)
#
#plotCat = SampledOPMData[cols]
#display(plotCat.head())
#print("plotCat Has {0} Records".format(len(plotCat)))
#print("Number of colums = ", len(cols))

%%time

cols = list(OPMDataMerged.select_dtypes(include=['object']))
dropCols = ["LOCTYP",
            "LOCTYPT",
            "OCCTYP",
            "OCCTYPT",
            "PPTYP",
            "PPTYPT",
            "AGYTYP",
            "OCCFAM",
            "PPGROUP",
            "PAYPLAN",
            "TOATYP",
            "WSTYP",
            "AGYSUBT",
            "AGELVL",
            "LOSLVL",
            "LOC",
            "OCC",
            "PATCO",
            "SALLVL",
            "TOA",
            "WORKSCH"]

for i in dropCols:
    cols.remove(i)

plotCat = OPMDataMerged[cols]
display(plotCat.head())
print("plotCat Has {0} Records".format(len(plotCat)))
print("Number of colums = ", len(cols))

plotCat Has 8170907 Records
Number of colums =  23
CPU times: user 3.01 s, sys: 1.33 s, total: 4.34 s
Wall time: 4.34 s

AGYSUB¶

High seperation among following:

Veterans Health Administration (VATA)
Forest Service (AG11)

GENDER¶

Similar separation distributions among males and females, except more terminations due to contract expiration among males

GSEGRD¶

High termination due to expired appt/other among following:

3
4
5

Bimodal Quit distribution with outlier spike at GSEGRD 9:

Distribution 1 from GSEGRD 3 to 8
Distribution 2 from GSEGRD 11 to 15

Individual transfers highest among levels 11, 12, 13

PPGRD¶

Majority of distribution resides in GS values per the GSEGRD observations described above.... Are other PPGRD values of any significance? What are corporate grades all about?

AGYT¶

Top three Agencies with separation:

AR-Department of the Army
AG-Department of Agriculture
VA-Department of Veteran Affairs

High contract termination in:

AG-Department of Agriculture
IN-Department of the Interior

While Veteran Affairs and Army both have many quits and many retirees, the Army has significantly more individual transfers (on par with retirements)

QTR¶

Most contract terminations in 1st and 4th quarters

Retirement peaks in 2nd quarter

Number of quits increases from one quarter to the next

*Bear in mind these are quarters from single year only so time-sensitive trends may not be applicable*

AGELVLT¶

High termination due to expired appt/other among following:

B
C

Number of Quits peaks at AGELVL D

Individual transfer counts mostly trend with Quits

Retirement highest at following:

I
J
K

LOSLVLT¶

Highest Quit count for LOSLVL A (< 1 year service) which then declines for levels B and C before spiking again at level D (5-9 years service)

Same pattern is observed for contract terminations but without any significant spikes with longer service

Large individual transfer spike at LOSLVL D (5-9 years service)

Retirement starts at LOSLVL D but trends upward to J

LOCT¶

Contract terminations comprise most California terminations among top total separation states

East Coast locations may possibly have most individual transfers, the most being in Washington DC

OCCFAMT¶

03xx-General Admin, clerical, and office svcs highest separation by far but indicates both high number of Quits and Retirements

Many quits in 06xx-Medical

04xx-Natural Resources again indicates high number of contract terminations

01xx-Social Science has even number of Quits and retirements

OCCT¶

PATCOT¶

PAYPLANT¶

Results skewed by GS

TOAT¶

WORKSCHT¶

Should model full time only

def subCountPlot(att1, att2, thresh):
    counts = plotCat.groupby([att1, att2]).size().unstack(fill_value=0) # Get att1 sizes by att2
    counts = pd.concat([counts,counts.sum(axis=1)], axis=1) # Calculate total for each att1 value and append total as new column
    counts.rename(columns={0:"Total"}, inplace=True)
    top = counts[counts["Total"] > thresh].index.tolist() # Obtain att1 values where total surpasses threshold
    
    zoom = plotCat[plotCat[att1].isin(top)] # Subset data to only the top att1 values
    f, (ax1, ax2) = plt.subplots(ncols=2, figsize=(20, 10), sharey=False)
    sns.countplot(y=att1, data=zoom, color="blue", ax=ax1); # Dark blue signifies zoomed data
    sns.countplot(y=att1, data=zoom, hue=att2, palette="hls", ax=ax2);

def percBarPlot(att1, att2, numColors):
    # Create count by att1 and att2
    counts = plotCat.groupby([att1, att2]).size().unstack(fill_value=0) # Get att1 sizes by att2
    counts = pd.concat([counts,counts.sum(axis=1)], axis=1) # Calculate total for each att1 value and append total as new column
    counts.rename(columns={0:"Total"}, inplace=True)
    #counts.drop('Total', axis=1).plot(kind='bar', stacked=True)
    
    # create cmap from sns color palette
    my_cmap = ListedColormap(sns.color_palette('hls', numColors).as_hex())

    # Create and plot percentage by att1 and att2
    nest1 = []
    for i in counts.values:
        nest2 = []
        for j in i:
            nest2.append(float(j/(i[len(i)-1:]))*100)
        nest1.append(nest2)
    perc = pd.DataFrame(nest1)
    perc = perc.set_index(counts.index.values)
    perc.columns = counts.columns
    perc.drop('Total', axis=1).plot(kind='bar', stacked=True, ylim=(0,100), figsize={13,6}, title=att1+' Percentage Plot', colormap=my_cmap)

temp = cols[:4] # for quick visualization debug only; may delete once complete

%%time

for i in cols:
    if i != 'SEP':
        plt.figure(i) # Required to create new figure each loop rather than drawing over previous object
        f, (ax1, ax2) = plt.subplots(ncols=2, figsize=(20, 10), sharey=False)
        sns.countplot(y=i, data=plotCat, color="lightblue", ax=ax1);
        sns.countplot(y=i, data=plotCat, hue="SEP", palette="hls", ax=ax2);
        
    if i == 'AGYSUB':
        subCountPlot(i, 'SEP', 10000)
    elif i == 'LOCT':
        subCountPlot(i, 'SEP', 4000)
    elif i == 'OCCT':
        subCountPlot(i, 'SEP', 2000)
    elif i == 'PPGRD':
        subCountPlot(i, 'SEP', 6000)
    elif i == 'AGYT':
        subCountPlot(i, 'SEP', 3000)

/usr/local/es7/lib/python3.5/site-packages/matplotlib/pyplot.py:524: RuntimeWarning: More than 20 figures have been opened. Figures created through the pyplot interface (`matplotlib.pyplot.figure`) are retained until explicitly closed and may consume too much memory. (To control this warning, see the rcParam `figure.max_open_warning`).
  max_open_warning, RuntimeWarning)

CPU times: user 10min 8s, sys: 13 s, total: 10min 21s
Wall time: 10min 20s

<matplotlib.figure.Figure at 0x7f72859fb780>

/usr/local/es7/lib/python3.5/site-packages/matplotlib/font_manager.py:1297: UserWarning: findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans
  (prop.get_family(), self.defaultFamily[fontext]))

<matplotlib.figure.Figure at 0x7f71ce323d68>

<matplotlib.figure.Figure at 0x7f71ccb748d0>

<matplotlib.figure.Figure at 0x7f71ce7facf8>

<matplotlib.figure.Figure at 0x7f71cc92d0f0>

<matplotlib.figure.Figure at 0x7f71afa42a20>

<matplotlib.figure.Figure at 0x7f71ae61d8d0>

%%time

for i in cols:
    if i != 'SEP':
        percBarPlot(i, 'SEP', len(plotCat.SEP.drop_duplicates()))

/usr/local/es7/lib/python3.5/site-packages/matplotlib/pyplot.py:524: RuntimeWarning: More than 20 figures have been opened. Figures created through the pyplot interface (`matplotlib.pyplot.figure`) are retained until explicitly closed and may consume too much memory. (To control this warning, see the rcParam `figure.max_open_warning`).
  max_open_warning, RuntimeWarning)

CPU times: user 1min 12s, sys: 3.72 s, total: 1min 15s
Wall time: 1min 15s

/usr/local/es7/lib/python3.5/site-packages/matplotlib/font_manager.py:1297: UserWarning: findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans
  (prop.get_family(), self.defaultFamily[fontext]))

percBarPlot('GSEGRD', 'SALLVLT', len(plotCat.SALLVLT.drop_duplicates()))

/usr/local/es7/lib/python3.5/site-packages/matplotlib/font_manager.py:1297: UserWarning: findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans
  (prop.get_family(), self.defaultFamily[fontext]))

percBarPlot('PATCOT', 'SALLVLT', len(plotCat.SALLVLT.drop_duplicates()))

/usr/local/es7/lib/python3.5/site-packages/matplotlib/font_manager.py:1297: UserWarning: findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans
  (prop.get_family(), self.defaultFamily[fontext]))

%%time

sns.set(style="whitegrid", palette="pastel", color_codes=True)

sns.violinplot(x="PATCOT", y="SALARY", hue="GENDER", data=OPMDataMerged[OPMDataMerged.GENDER != 'Z'], split=True,
               inner="quart", palette={"M": "b", "F": "pink"})
sns.despine(left=True)

CPU times: user 22.5 s, sys: 1min 4s, total: 1min 26s
Wall time: 14 s

/usr/local/es7/lib/python3.5/site-packages/matplotlib/font_manager.py:1297: UserWarning: findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans
  (prop.get_family(), self.defaultFamily[fontext]))

%%time

# Draw a nested violinplot and split the violins for easier comparison
sns.violinplot(x="SEP", y="SALARY", hue="GENDER", data=OPMDataMerged[OPMDataMerged.GENDER != 'Z'], split=True,
               inner="quart", palette={"M": "b", "F": "pink"})
sns.despine(left=True)

CPU times: user 17.1 s, sys: 38.5 s, total: 55.6 s
Wall time: 12.4 s

/usr/local/es7/lib/python3.5/site-packages/matplotlib/font_manager.py:1297: UserWarning: findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans
  (prop.get_family(), self.defaultFamily[fontext]))

%%time

sns.factorplot(x="SEP", y="SALARY", hue="GENDER", col="PATCOT",
               data=OPMDataMerged[OPMDataMerged.GENDER != 'Z'],
               kind="violin", split=True, aspect=.4, size=10);

/usr/local/es7/lib/python3.5/site-packages/matplotlib/font_manager.py:1297: UserWarning: findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans
  (prop.get_family(), self.defaultFamily[fontext]))

CPU times: user 35.3 s, sys: 16.7 s, total: 52 s
Wall time: 38.1 s

<seaborn.axisgrid.FacetGrid at 0x7f71a1bebfd0>

#%%time
#
#sns.factorplot(x="SEP", y="SALARY", col="PATCOT", data=OPMDataMerged,
#               kind="violin", split=True, aspect=.4, size=10, palette = "hls");

#%%time
#
#g = sns.PairGrid(data=OPMDataMerged,
#                 x_vars=["SEP","PATCOT"],
#                 y_vars=["SALARY", "LOS", "LowerLimitAge", "YearsToRetirement"],
#                 aspect=1, size=10)
#g.map(sns.violinplot, palette="pastel");

del(plotNumeric, plotCat)

Focusing in on our Target Demographic¶

After analyzing the above plots for our categorical data, we have decided to narrow our focus due to the large variability in the dataset. We take the below actions on our dataset:

Keep only Full-time Nonseasonal observations
Remove the location US-SUPPRESSED (SEE DATA DEFINITIONS) due to apparent bias towards unknowns in the non-separation data
Keep only General Schedele Grades above 7.
Focus model generation on White Collar Jobs only
Create a Training set for the Professional PATCO value, and a Testing set for Administration

In addition, we have opted to remove the below attributes for model generation:

Datecode, QTR; Although very relevant for merging data from alternate sources, we do not have several years of data so this does not bring us much value
All Agency Attributes(AGYTYP,AGYTYPT,AGY,AGYT,AGYSUB,AGYSUBT); We are not concerned with agencies
Gender; Missing values for Non-Separation observations
Count; Missing values for Non-Separation observations; Also, all values = 1 so not very useful
PAYPLAN,PAYPLANT,PPGRD; Much too granular than we care for
LOSLVL,LOSLVLT; we have a numerical version of this attribute
OCC,OCCT; Much too granular than we care for

Our goal is to limit our focus to Professional occupations, build a model, then test that generated model on the Administration segment of the population.

%%time
print(len(OPMDataMerged))
#Removing Attributes
cols = list(OPMDataMerged.columns)
dropCols = ["QTR",
            "AGYTYP",
            "AGYTYPT",
            "AGY",
            "AGYT",
            "AGYSUB",
            "AGYSUBT",
            "GENDER",
            "COUNT",
            "PAYPLAN",
            "PAYPLANT",
            "PPGRD",
            "LOSLVL",
            "LOSLVLT",
            "SALLVL",
            "SALLVLT",
            "OCC",
            "OCCT"]

for i in dropCols:
    if i in cols:
        cols.remove(i)

OPMDataMerged = OPMDataMerged[cols]

# Keep only Full-time Nonseasonal observations
OPMDataMerged = OPMDataMerged[OPMDataMerged["WORKSCH"] == "F"]

#Remove the location US-SUPPRESSED (SEE DATA DEFINITIONS)
OPMDataMerged = OPMDataMerged[OPMDataMerged["LOC"] != "US"]

#Keep only General Schedele Grades above 7.
OPMDataMerged["GSEGRD"] = OPMDataMerged["GSEGRD"].apply(pd.to_numeric)
OPMDataMerged = OPMDataMerged[OPMDataMerged["GSEGRD"] >= 7]

#Focus model generation on White Collar Jobs only
OPMDataMerged = OPMDataMerged[OPMDataMerged["OCCTYP"] == "1"]

#Create a Training set for the Professional PATCO value, and a Testing set for Administration
OPMDataMergedProf = OPMDataMerged[OPMDataMerged["PATCO"] == "1"]
OPMDataMergedAdmin = OPMDataMerged[OPMDataMerged["PATCO"] == "2"]

8170907
CPU times: user 2min 1s, sys: 3.49 s, total: 2min 5s
Wall time: 2min 5s

display(OPMDataMergedProf.head())
print(len(OPMDataMergedProf))

1283383

display(OPMDataMergedAdmin.head())
print(len(OPMDataMergedAdmin))

2131840

Sampling¶

#curious on stratum SEP counts for full remaining data
stratum = pd.DataFrame({'StratCount' : OPMDataMerged.groupby(["SEP"]).size()}).reset_index()

display(stratum)

#Assess Stratum SEP Counts for Prof, for use in sampling
maxSize=7500
stratumProf = pd.DataFrame({'StratCount' : OPMDataMergedProf.groupby(["SEP"]).size()}).reset_index()

stratumProf.loc[stratumProf["StratCount"]>maxSize,"StratCountSample"] = maxSize
stratumProf.loc[stratumProf["StratCount"]<=maxSize,"StratCountSample"] = stratumProf["StratCount"]
#else: stratum["StratCountSample"] = stratum["StratCount"]

display(stratumProf)

#Assess Stratum SEP Counts for Admin, for use in sampling
maxSize=7500
stratumAdmin = pd.DataFrame({'StratCount' : OPMDataMergedAdmin.groupby(["SEP"]).size()}).reset_index()

stratumAdmin.loc[stratumAdmin["StratCount"]>maxSize,"StratCountSample"] = maxSize
stratumAdmin.loc[stratumAdmin["StratCount"]<=maxSize,"StratCountSample"] = stratumAdmin["StratCount"]
#else: stratum["StratCountSample"] = stratum["StratCount"]

display(stratumAdmin)

%%time
def aggStratPop(stratum, OPMDataMerged):
    AggStrat = []

    for i in range(0,len(stratum)):
        sep = stratum["SEP"].ix[i]
        StratCountSample = stratum["StratCountSample"].ix[i]
        print("Stratum Sample Size Calculations for SEP: {}".format(sep))   
        AggStrat.append(pd.DataFrame({'StratCount' : OPMDataMerged[OPMDataMerged["SEP"]==sep].groupby(["DATECODE", "AGELVL"]).size()}).reset_index())
        AggStrat[i]["SEP"] = sep
        AggStrat[i]["TotalCount"] = len(OPMDataMerged[OPMDataMerged["SEP"]==sep])
        AggStrat[i]["p"] = AggStrat[i]["StratCount"] / AggStrat[i]["TotalCount"]
        AggStrat[i]["StratCountSample"] = StratCountSample
        AggStrat[i]["StratSampleSize"] = round(AggStrat[i]["p"] * StratCountSample).apply(int)

        display(AggStrat[i].head())
        print("totalStratumSampleSize: ", AggStrat[i]["StratSampleSize"].sum())
        #print(len(AggStrat[i]))
    return AggStrat

CPU times: user 5 µs, sys: 1e+03 ns, total: 6 µs
Wall time: 10 µs

def SampleStrata(stratum, OPMDataMerged, FileName):
    AggStrat = aggStratPop(stratum, OPMDataMerged)

    SampledOPMStratumDataList = []

    for i,StratSampleSize in enumerate(AggStrat):
        SampledOPMStratumData = []
        for j in range(0,len(StratSampleSize)):
            SEP = StratSampleSize["SEP"].ix[j]
            DATECODE = StratSampleSize["DATECODE"].ix[j]
            AGELVL = StratSampleSize["AGELVL"].ix[j]
            SampleSize = StratSampleSize["StratSampleSize"].ix[j]
            print(SEP, DATECODE, AGELVL, SampleSize)

            SampledOPMStratumDataList.append(OPMDataMerged[(OPMDataMerged["SEP"]==SEP) 
                                                    & (OPMDataMerged["DATECODE"]==DATECODE) 
                                                    & (OPMDataMerged["AGELVL"]==AGELVL)].sample(SampleSize,  random_state=SampleSize))
        SampledOPMStratumData.append(pd.concat(SampledOPMStratumDataList))
        clear_display()
    SampledOPMData = pd.concat(SampledOPMStratumData).reset_index()
    del SampledOPMData["index"]
    pickleObject(SampledOPMData, FileName)
    clear_display()

    return SampledOPMData

Using a seed value equal to each strata sample size, we take random samples according to the computed sizes above. We loop through each Separation Type's Aggregated Strata Sample Sizes; Identify all observations matching on Datecode, Separation Type, and AgeLevel; and finally sample those observations with the computed sample size.

%%time
##Prof Data Sampling
if os.path.isfile(PickleJarPath+"/SampledOPMDataProf.pkl"):
    print("Found the File! Loading Pickle Now!")
    SampledOPMDataProf = unpickleObject("SampledOPMDataProf")
else:
    SampledOPMDataProf= SampleStrata(stratumProf, OPMDataMergedProf, "SampledOPMDataProf")

Found the File! Loading Pickle Now!
CPU times: user 26.3 ms, sys: 3.21 ms, total: 29.5 ms
Wall time: 29.3 ms

%%time
print(len(SampledOPMDataProf))
display(SampledOPMDataProf.head())
display(pd.DataFrame({'StratCount' : SampledOPMDataProf.groupby(["SEP"]).size()}).reset_index())

30227

CPU times: user 52.8 ms, sys: 538 µs, total: 53.4 ms
Wall time: 48 ms

%%time
#### Analyze Missing Values
filtered_msnoData = msno.nullity_sort(msno.nullity_filter(SampledOPMDataProf, filter='bottom', n=15, p=0.999), sort='descending')
msno.matrix(filtered_msnoData)

del filtered_msnoData

/usr/local/es7/lib/python3.5/site-packages/matplotlib/axes/_base.py:2903: UserWarning: Attempting to set identical left==right results
in singular transformations; automatically expanding.
left=-0.5, right=-0.5
  'left=%s, right=%s') % (left, right))
/usr/local/es7/lib/python3.5/site-packages/matplotlib/font_manager.py:1297: UserWarning: findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans
  (prop.get_family(), self.defaultFamily[fontext]))

CPU times: user 333 ms, sys: 248 ms, total: 581 ms
Wall time: 296 ms

%%time
##Admin Data Sampling
if os.path.isfile(PickleJarPath+"/SampledOPMDataAdmin.pkl"):
    print("Found the File! Loading Pickle Now!")
    SampledOPMDataAdmin = unpickleObject("SampledOPMDataAdmin")
else:
    SampledOPMDataAdmin= SampleStrata(stratumAdmin, OPMDataMergedAdmin, "SampledOPMDataAdmin")

Found the File! Loading Pickle Now!
CPU times: user 43 ms, sys: 62 ms, total: 105 ms
Wall time: 34.7 ms

%%time
print(len(SampledOPMDataAdmin))
display(SampledOPMDataAdmin.head())
display(pd.DataFrame({'StratCount' : SampledOPMDataAdmin.groupby(["SEP"]).size()}).reset_index())

36956

CPU times: user 57.5 ms, sys: 17.1 ms, total: 74.6 ms
Wall time: 49.5 ms

%%time
#### Analyze Missing Values
filtered_msnoData = msno.nullity_sort(msno.nullity_filter(SampledOPMDataAdmin, filter='bottom', n=15, p=0.999), sort='descending')
msno.matrix(filtered_msnoData)

del filtered_msnoData

/usr/local/es7/lib/python3.5/site-packages/matplotlib/axes/_base.py:2903: UserWarning: Attempting to set identical left==right results
in singular transformations; automatically expanding.
left=-0.5, right=-0.5
  'left=%s, right=%s') % (left, right))
/usr/local/es7/lib/python3.5/site-packages/matplotlib/font_manager.py:1297: UserWarning: findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans
  (prop.get_family(), self.defaultFamily[fontext]))

CPU times: user 351 ms, sys: 232 ms, total: 583 ms
Wall time: 298 ms

%%time
## Describe Summary for our Model Professional Subgroup for Modeling
display(SampledOPMDataProf.describe().transpose())

CPU times: user 99.1 ms, sys: 83.9 ms, total: 183 ms
Wall time: 76.8 ms

#%%time

#OPMDataMerged.to_csv("OPMDataMerged.csv")

#os.path.getsize("OPMDataMerged.csv") #Display file size in bytes

Review Visualizations post-Data removal and sampling¶

Chris... can you use the SampledOPMDataProf dataset, and re-run the Visuals?

%%time


cols = list(SampledOPMDataProf.select_dtypes(include=['float64', 'int64']))
cols.remove('BLS_FEDERAL_OtherSep_Rate')
cols.remove('BLS_FEDERAL_Quits_Rate')
cols.remove('BLS_FEDERAL_TotalSep_Level')
cols.remove('BLS_FEDERAL_JobOpenings_Rate')
cols.remove('BLS_FEDERAL_OtherSep_Level')
cols.remove('BLS_FEDERAL_Quits_Level')
cols.remove('BLS_FEDERAL_JobOpenings_Level')
cols.remove('BLS_FEDERAL_Layoffs_Rate')
cols.remove('BLS_FEDERAL_Layoffs_Level')
cols.remove('BLS_FEDERAL_TotalSep_Rate')
cols.append('SEP')
display(cols)

plotNumeric = SampledOPMDataProf[cols]

# Create binary separation attribute for EDA correlation review
#plotNumeric["SEP_bin"] = plotNumeric.SEP.replace("NS", 1)
#plotNumeric.loc[plotNumeric['SEP_bin'] != 1, 'SEP_bin'] = 0
#plotNumeric.SEP_bin = plotNumeric.SEP_bin.apply(pd.to_numeric)
AttSplit = pd.get_dummies(plotNumeric['SEP'],prefix='SEP')
display(AttSplit.head())
plotNumeric = pd.concat((plotNumeric,AttSplit),axis=1) # add back into the dataframe

display(plotNumeric.head())
print("plotNumeric has {0} Records".format(len(plotNumeric)))
#print(plotNumeric.SEP_bin.dtype)

['GSEGRD',
 'SALARY',
 'LOS',
 'SEPCount_EFDATE_OCC',
 'SEPCount_EFDATE_LOC',
 'IndAvgSalary',
 'SalaryOverUnderIndAvg',
 'LowerLimitAge',
 'YearsToRetirement',
 'SALARYLog',
 'LOSSqrt',
 'SEPCount_EFDATE_OCCLog',
 'SEPCount_EFDATE_LOCLog',
 'IndAvgSalaryLog',
 'SEP']

plotNumeric has 30227 Records
CPU times: user 38.2 ms, sys: 1.24 ms, total: 39.4 ms
Wall time: 36.8 ms

%%time

sns.set(font_scale=1)
sns.pairplot(plotNumeric.drop(["SEP_NS", 
                               "SEP_SA", 
                               "SEP_SB", 
                               "SEP_SC", 
                               "SEP_SD", 
                               "SEP_SE", 
                               "SEP_SF", 
                               "SEP_SG", 
                               "SEP_SH", 
                               "SEP_SI", 
                               "SEP_SJ", 
                               "SEP_SK", 
                               "SEP_SL"
                              ], axis=1), hue = 'SEP', palette="hls", plot_kws={"s": 50})

/usr/local/es7/lib/python3.5/site-packages/matplotlib/font_manager.py:1297: UserWarning: findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans
  (prop.get_family(), self.defaultFamily[fontext]))

CPU times: user 1min 32s, sys: 1min 29s, total: 3min 1s
Wall time: 1min 14s

%%time

# Function modified from https://stackoverflow.com/questions/29530355/plotting-multiple-histograms-in-grid
sns.set()

def draw_histograms(df, variables, n_rows, n_cols):
    fig=plt.figure(figsize=(20,20))
    for i, var_name in enumerate(variables):
        ax=fig.add_subplot(n_rows,n_cols,i+1)
        df[var_name].hist(bins=20,ax=ax, color='#58D68D')
        ax.set_title(var_name+" Distribution")
    fig.tight_layout()  # Improves appearance a bit.
    plt.show()

draw_histograms(plotNumeric.drop(['SEP',
                                  "SEP_NS", 
                               "SEP_SA", 
                               "SEP_SB", 
                               "SEP_SC", 
                               "SEP_SD", 
                               "SEP_SE", 
                               "SEP_SF", 
                               "SEP_SG", 
                               "SEP_SH", 
                               "SEP_SI", 
                               "SEP_SJ", 
                               "SEP_SK", 
                               "SEP_SL"
                              ], axis=1),
                plotNumeric.drop(['SEP',
                                  "SEP_NS", 
                               "SEP_SA", 
                               "SEP_SB", 
                               "SEP_SC", 
                               "SEP_SD", 
                               "SEP_SE", 
                               "SEP_SF", 
                               "SEP_SG", 
                               "SEP_SH", 
                               "SEP_SI", 
                               "SEP_SJ", 
                               "SEP_SK", 
                               "SEP_SL"
                              ], axis=1).columns, 6, 3)

/usr/local/es7/lib/python3.5/site-packages/matplotlib/font_manager.py:1297: UserWarning: findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans
  (prop.get_family(), self.defaultFamily[fontext]))

CPU times: user 3.45 s, sys: 2.77 s, total: 6.22 s
Wall time: 2.88 s

%%time
# Inspired by http://seaborn.pydata.org/examples/many_pairwise_correlations.html

#plt.matshow(plotNumeric.corr())

sns.set(style='white')
corr = plotNumeric.drop(['SEP'], axis=1).corr()

# Generate a mask for the upper triangle
mask = np.zeros_like(corr, dtype=np.bool)
mask[np.triu_indices_from(mask, k=1)] = True

# Set up the matplotlib figure
f, ax = plt.subplots(figsize=(20, 20))

# Generate a custom diverging colormap
cmap = sns.diverging_palette(250, 10, as_cmap=True)

# Draw the heatmap with the mask and correct aspect ratio
sns.set(font_scale=0.95)
heatCorr = sns.heatmap(corr, mask=mask, cmap=cmap, vmax=1, vmin=-1,
                       square=True, annot=True, linewidths=1,
                       cbar_kws={"shrink": .5}, ax=ax, fmt='.1g')
#heatCorr.
ax.tick_params(labelsize=15)
cax = plt.gcf().axes[-1]
cax.tick_params(labelsize=15)

sns.plt.show()
#sns.heatmap(corr, annot=True, linewidths=0.01, cmap=cmap, ax=ax)

/usr/local/es7/lib/python3.5/site-packages/matplotlib/font_manager.py:1297: UserWarning: findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans
  (prop.get_family(), self.defaultFamily[fontext]))

CPU times: user 3.37 s, sys: 925 ms, total: 4.29 s
Wall time: 3.24 s

%%time

cols = list(SampledOPMDataProf.select_dtypes(include=['object']))
dropCols = ["LOCTYP",
            "LOCTYPT",
            "OCCTYP",
            "OCCTYPT",
            "PPTYP",
            "PPTYPT",
            "AGYTYP",
            "OCCFAM",
            "PPGROUP",
            "PAYPLAN",
            "TOATYP",
            "WSTYP",
            "AGYSUBT",
            "AGELVL",
            "LOSLVL",
            "LOC",
            "OCC",
            "PATCO",
            "SALLVL",
            "TOA",
            "WORKSCH"]

for i in dropCols:
    if(i in list(SampledOPMDataProf.columns)): cols.remove(i)

plotCat = SampledOPMDataProf[cols]
display(plotCat.head())
print("plotCat Has {0} Records".format(len(plotCat)))
print("Number of colums = ", len(cols))

plotCat Has 30227 Records
Number of colums =  11
CPU times: user 19.7 ms, sys: 2.93 ms, total: 22.7 ms
Wall time: 20.6 ms

%%time

for i in cols:
    if i != 'SEP':
        plt.figure(i) # Required to create new figure each loop rather than drawing over previous object
        f, (ax1, ax2) = plt.subplots(ncols=2, figsize=(20, 10), sharey=False)
        sns.countplot(y=i, data=plotCat, color="lightblue", ax=ax1);
        sns.countplot(y=i, data=plotCat, hue="SEP", palette="hls", ax=ax2);
        
    if i == 'AGYSUB':
        subCountPlot(i, 'SEP', 10000)
    elif i == 'LOCT':
        subCountPlot(i, 'SEP', 1000)
    elif i == 'OCCT':
        subCountPlot(i, 'SEP', 2000)
    elif i == 'PPGRD':
        subCountPlot(i, 'SEP', 6000)
    elif i == 'AGYT':
        subCountPlot(i, 'SEP', 3000)

CPU times: user 5.05 s, sys: 30 ms, total: 5.08 s
Wall time: 5.05 s

/usr/local/es7/lib/python3.5/site-packages/matplotlib/pyplot.py:524: RuntimeWarning: More than 20 figures have been opened. Figures created through the pyplot interface (`matplotlib.pyplot.figure`) are retained until explicitly closed and may consume too much memory. (To control this warning, see the rcParam `figure.max_open_warning`).
  max_open_warning, RuntimeWarning)

<matplotlib.figure.Figure at 0x7f71b4d40978>

/usr/local/es7/lib/python3.5/site-packages/matplotlib/font_manager.py:1297: UserWarning: findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans
  (prop.get_family(), self.defaultFamily[fontext]))

<matplotlib.figure.Figure at 0x7f7198b20518>

<matplotlib.figure.Figure at 0x7f720c416320>

<matplotlib.figure.Figure at 0x7f71a8343630>

<matplotlib.figure.Figure at 0x7f719cc031d0>

<matplotlib.figure.Figure at 0x7f719ea51dd8>

<matplotlib.figure.Figure at 0x7f719dc570b8>

%%time

for i in cols:
    if i != 'SEP':
        percBarPlot(i, 'SEP', len(plotCat.SEP.drop_duplicates()))

CPU times: user 2.47 s, sys: 34.9 ms, total: 2.5 s
Wall time: 2.48 s

/usr/local/es7/lib/python3.5/site-packages/matplotlib/font_manager.py:1297: UserWarning: findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans
  (prop.get_family(), self.defaultFamily[fontext]))

%%time

sns.set(style="whitegrid", palette="pastel", color_codes=True)

sns.violinplot(x="PATCOT", y="SALARY", data=SampledOPMDataProf, split=True,
               inner="quart")
sns.despine(left=True)

CPU times: user 1.66 s, sys: 8.85 s, total: 10.5 s
Wall time: 254 ms

/usr/local/es7/lib/python3.5/site-packages/matplotlib/font_manager.py:1297: UserWarning: findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans
  (prop.get_family(), self.defaultFamily[fontext]))

%%time

# Draw a nested violinplot and split the violins for easier comparison
sns.violinplot(x="SEP", y="SALARY", data=SampledOPMDataProf, split=True,
               inner="box", scale="area", cut=0)
sns.despine(left=True)

CPU times: user 341 ms, sys: 248 ms, total: 589 ms
Wall time: 295 ms

/usr/local/es7/lib/python3.5/site-packages/matplotlib/font_manager.py:1297: UserWarning: findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans
  (prop.get_family(), self.defaultFamily[fontext]))

#%%time
#
#sns.factorplot(x="SEP", y="SALARY", col="PATCOT",
#               data=SampledOPMDataProf,
#               kind="violin", split=True, aspect=.5, size=15);

#%%time
#
#sns.factorplot(x="SEP", y="SALARY", col="PATCOT", data=SampledOPMDataProf,
#               kind="violin", split=True, aspect=.4, size=10);

%%time

g = sns.PairGrid(data=SampledOPMDataProf,
                 x_vars=["SEP","PATCOT"],
                 y_vars=["SALARY", "LOS", "LowerLimitAge", "YearsToRetirement"],
                 aspect=1, size=10)
g.map(sns.violinplot, palette="pastel", inner="quart");

/usr/local/es7/lib/python3.5/site-packages/matplotlib/font_manager.py:1297: UserWarning: findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans
  (prop.get_family(), self.defaultFamily[fontext]))

CPU times: user 7.67 s, sys: 33.2 s, total: 40.9 s
Wall time: 2.52 s

There are several separation types we would like to either roll up, or remove altogether.

We have chosen to roll separation into two Binary Categories.

1) NS, Non-Separation comprised of: a) NS, Non-Separation b) SA,Transfer Out - Individual Transfer c) SB,Transfer Out - Mass Transfer d) SD,Retirement - Voluntary e) SE,Retirement - Early Out f) SF,Retirement - Disability g) SG,Retirement - Other

2) SC, Quit

SampledOPMDataProf = SampledOPMDataProf[((SampledOPMDataProf["SEP"] == "NS") | (SampledOPMDataProf["SEP"] == "SD") | (SampledOPMDataProf["SEP"] == "SE") | (SampledOPMDataProf["SEP"] == "SF") | (SampledOPMDataProf["SEP"] == "SH") | (SampledOPMDataProf["SEP"] == "SA") | (SampledOPMDataProf["SEP"] == "SB") | (SampledOPMDataProf["SEP"] == "SC"))]

SampledOPMDataProf.loc[(SampledOPMDataProf["SEP"] != "SC") , "SEP"]="NS"

SampledOPMDataAdmin = SampledOPMDataAdmin[((SampledOPMDataAdmin["SEP"] == "NS") | (SampledOPMDataAdmin["SEP"] == "SD") | (SampledOPMDataAdmin["SEP"] == "SE") | (SampledOPMDataAdmin["SEP"] == "SF") | (SampledOPMDataAdmin["SEP"] == "SH") | (SampledOPMDataAdmin["SEP"] == "SA") | (SampledOPMDataAdmin["SEP"] == "SB") | (SampledOPMDataAdmin["SEP"] == "SC"))]

SampledOPMDataAdmin.loc[(SampledOPMDataAdmin["SEP"] != "SC") , "SEP"]="NS"

#Assess Stratum SEP Counts for Prof, for use in sampling
maxSize=7500
stratumProf = pd.DataFrame({'StratCount' : SampledOPMDataProf.groupby(["SEP"]).size()}).reset_index()

stratumProf.loc[stratumProf["StratCount"]>maxSize,"StratCountSample"] = maxSize
stratumProf.loc[stratumProf["StratCount"]<=maxSize,"StratCountSample"] = stratumProf["StratCount"]
#else: stratum["StratCountSample"] = stratum["StratCount"]

display(stratumProf)

#Assess Stratum SEP Counts for Admin, for use in sampling
maxSize=7500
stratumAdmin = pd.DataFrame({'StratCount' : SampledOPMDataAdmin.groupby(["SEP"]).size()}).reset_index()

stratumAdmin.loc[stratumAdmin["StratCount"]>maxSize,"StratCountSample"] = maxSize
stratumAdmin.loc[stratumAdmin["StratCount"]<=maxSize,"StratCountSample"] = stratumAdmin["StratCount"]
#else: stratum["StratCountSample"] = stratum["StratCount"]

display(stratumAdmin)

SampledOPMDataProf= SampleStrata(stratumProf, SampledOPMDataProf, "SampledOPMDataProfBinary")
SampledOPMDataAdmin= SampleStrata(stratumProf, SampledOPMDataAdmin, "SampledOPMDataProfBinary")

Stratum Sample Size Calculations for SEP: NS

totalStratumSampleSize:  7497
Stratum Sample Size Calculations for SEP: SC

totalStratumSampleSize:  7423
NS 201410 B 1
NS 201410 C 20
NS 201410 D 30
NS 201410 E 26
NS 201410 F 26
NS 201410 G 27
NS 201410 H 25
NS 201410 I 66
NS 201410 J 77
NS 201410 K 68
NS 201411 B 1
NS 201411 C 25
NS 201411 D 42
NS 201411 E 35
NS 201411 F 34
NS 201411 G 26
NS 201411 H 34
NS 201411 I 59
NS 201411 J 56
NS 201411 K 42
NS 201412 B 5
NS 201412 C 53
NS 201412 D 99
NS 201412 E 95
NS 201412 F 90
NS 201412 G 99
NS 201412 H 134
NS 201412 I 192
NS 201412 J 200
NS 201412 K 169
NS 201501 B 1
NS 201501 C 12
NS 201501 D 27
NS 201501 E 30
NS 201501 F 19
NS 201501 G 26
NS 201501 H 32
NS 201501 I 182
NS 201501 J 218
NS 201501 K 190
NS 201502 B 1
NS 201502 C 16
NS 201502 D 23
NS 201502 E 18
NS 201502 F 17
NS 201502 G 16
NS 201502 H 27
NS 201502 I 51
NS 201502 J 67
NS 201502 K 49
NS 201503 B 6
NS 201503 C 57
NS 201503 D 117
NS 201503 E 108
NS 201503 F 105
NS 201503 G 118
NS 201503 H 140
NS 201503 I 142
NS 201503 J 133
NS 201503 K 81
NS 201504 C 14
NS 201504 D 27
NS 201504 E 29
NS 201504 F 25
NS 201504 G 20
NS 201504 H 31
NS 201504 I 54
NS 201504 J 67
NS 201504 K 64
NS 201505 B 1
NS 201505 C 23
NS 201505 D 41
NS 201505 E 39
NS 201505 F 35
NS 201505 G 37
NS 201505 H 40
NS 201505 I 105
NS 201505 J 112
NS 201505 K 93
NS 201506 B 8
NS 201506 C 59
NS 201506 D 114
NS 201506 E 107
NS 201506 F 96
NS 201506 G 102
NS 201506 H 137
NS 201506 I 159
NS 201506 J 149
NS 201506 K 94
NS 201507 B 1
NS 201507 C 18
NS 201507 D 29
NS 201507 E 33
NS 201507 F 23
NS 201507 G 21
NS 201507 H 28
NS 201507 I 80
NS 201507 J 89
NS 201507 K 78
NS 201508 B 0
NS 201508 C 18
NS 201508 D 34
NS 201508 E 22
NS 201508 F 21
NS 201508 G 22
NS 201508 H 30
NS 201508 I 54
NS 201508 J 70
NS 201508 K 50
NS 201509 B 7
NS 201509 C 57
NS 201509 D 121
NS 201509 E 111
NS 201509 F 99
NS 201509 G 110
NS 201509 H 134
NS 201509 I 151
NS 201509 J 130
NS 201509 K 89
SC 201410 B 10
SC 201410 C 92
SC 201410 D 154
SC 201410 E 118
SC 201410 F 80
SC 201410 G 61
SC 201410 H 64
SC 201410 I 48
SC 201410 J 24
SC 201410 K 11
SC 201411 B 6
SC 201411 C 72
SC 201411 D 90
SC 201411 E 85
SC 201411 F 70
SC 201411 G 55
SC 201411 H 61
SC 201411 I 40
SC 201411 J 17
SC 201411 K 7
SC 201412 B 3
SC 201412 C 63
SC 201412 D 103
SC 201412 E 81
SC 201412 F 66
SC 201412 G 45
SC 201412 H 47
SC 201412 I 43
SC 201412 J 12
SC 201412 K 14
SC 201501 B 4
SC 201501 C 106
SC 201501 D 132
SC 201501 E 122
SC 201501 F 95
SC 201501 G 71
SC 201501 H 72
SC 201501 I 42
SC 201501 J 23
SC 201501 K 9
SC 201502 B 6
SC 201502 C 63
SC 201502 D 98
SC 201502 E 79
SC 201502 F 71
SC 201502 G 49
SC 201502 H 50
SC 201502 I 44
SC 201502 J 16
SC 201502 K 12
SC 201503 B 9
SC 201503 C 60
SC 201503 D 101
SC 201503 E 91
SC 201503 F 52
SC 201503 G 56
SC 201503 H 38
SC 201503 I 38
SC 201503 J 21
SC 201503 K 11
SC 201504 B 11
SC 201504 C 86
SC 201504 D 104
SC 201504 E 72
SC 201504 F 58
SC 201504 G 58
SC 201504 H 56
SC 201504 I 36
SC 201504 J 13
SC 201504 K 12
SC 201505 B 19
SC 201505 C 140
SC 201505 D 162
SC 201505 E 105
SC 201505 F 105
SC 201505 G 72
SC 201505 H 66
SC 201505 I 56
SC 201505 J 28
SC 201505 K 11
SC 201506 B 16
SC 201506 C 109
SC 201506 D 139
SC 201506 E 126
SC 201506 F 84
SC 201506 G 76
SC 201506 H 68
SC 201506 I 45
SC 201506 J 21
SC 201506 K 16
SC 201507 B 24
SC 201507 C 139
SC 201507 D 141
SC 201507 E 120
SC 201507 F 103
SC 201507 G 71
SC 201507 H 57
SC 201507 I 56
SC 201507 J 29
SC 201507 K 15
SC 201508 B 17
SC 201508 C 105
SC 201508 D 164
SC 201508 E 140
SC 201508 F 91
SC 201508 G 82
SC 201508 H 65
SC 201508 I 48
SC 201508 J 15
SC 201508 K 6
SC 201509 B 12
SC 201509 C 102
SC 201509 D 160
SC 201509 E 97
SC 201509 F 87
SC 201509 G 69
SC 201509 H 66
SC 201509 I 45
SC 201509 J 29
SC 201509 K 15
Stratum Sample Size Calculations for SEP: NS

totalStratumSampleSize:  7495
Stratum Sample Size Calculations for SEP: SC

totalStratumSampleSize:  7423
NS 201410 B 1
NS 201410 C 18
NS 201410 D 41
NS 201410 E 30
NS 201410 F 35
NS 201410 G 42
NS 201410 H 53
NS 201410 I 75
NS 201410 J 66
NS 201410 K 36
NS 201411 B 2
NS 201411 C 19
NS 201411 D 42
NS 201411 E 39
NS 201411 F 46
NS 201411 G 45
NS 201411 H 52
NS 201411 I 70
NS 201411 J 48
NS 201411 K 31
NS 201412 B 2
NS 201412 C 32
NS 201412 D 71
NS 201412 E 79
NS 201412 F 86
NS 201412 G 127
NS 201412 H 174
NS 201412 I 207
NS 201412 J 189
NS 201412 K 128
NS 201501 B 1
NS 201501 C 14
NS 201501 D 24
NS 201501 E 30
NS 201501 F 35
NS 201501 G 35
NS 201501 H 51
NS 201501 I 189
NS 201501 J 191
NS 201501 K 121
NS 201502 B 2
NS 201502 C 18
NS 201502 D 28
NS 201502 E 25
NS 201502 F 27
NS 201502 G 36
NS 201502 H 37
NS 201502 I 58
NS 201502 J 52
NS 201502 K 30
NS 201503 B 2
NS 201503 C 38
NS 201503 D 88
NS 201503 E 92
NS 201503 F 91
NS 201503 G 132
NS 201503 H 173
NS 201503 I 156
NS 201503 J 110
NS 201503 K 61
NS 201504 B 1
NS 201504 C 13
NS 201504 D 30
NS 201504 E 28
NS 201504 F 33
NS 201504 G 31
NS 201504 H 45
NS 201504 I 71
NS 201504 J 61
NS 201504 K 39
NS 201505 B 0
NS 201505 C 21
NS 201505 D 50
NS 201505 E 43
NS 201505 F 40
NS 201505 G 47
NS 201505 H 69
NS 201505 I 112
NS 201505 J 93
NS 201505 K 63
NS 201506 B 2
NS 201506 C 35
NS 201506 D 79
NS 201506 E 86
NS 201506 F 90
NS 201506 G 125
NS 201506 H 163
NS 201506 I 179
NS 201506 J 121
NS 201506 K 68
NS 201507 B 0
NS 201507 C 12
NS 201507 D 42
NS 201507 E 31
NS 201507 F 33
NS 201507 G 42
NS 201507 H 67
NS 201507 I 101
NS 201507 J 85
NS 201507 K 50
NS 201508 B 1
NS 201508 C 11
NS 201508 D 33
NS 201508 E 33
NS 201508 F 26
NS 201508 G 40
NS 201508 H 53
NS 201508 I 67
NS 201508 J 56
NS 201508 K 33
NS 201509 B 3
NS 201509 C 40
NS 201509 D 89
NS 201509 E 103
NS 201509 F 98
NS 201509 G 138
NS 201509 H 176
NS 201509 I 176
NS 201509 J 126
NS 201509 K 59
SC 201410 B 16
SC 201410 C 76
SC 201410 D 140
SC 201410 E 109
SC 201410 F 65
SC 201410 G 91
SC 201410 H 88
SC 201410 I 46
SC 201410 J 25
SC 201410 K 8
SC 201411 B 7
SC 201411 C 66
SC 201411 D 94
SC 201411 E 90
SC 201411 F 80
SC 201411 G 75
SC 201411 H 70
SC 201411 I 38
SC 201411 J 11
SC 201411 K 7
SC 201412 B 2
SC 201412 C 59
SC 201412 D 92
SC 201412 E 67
SC 201412 F 56
SC 201412 G 61
SC 201412 H 65
SC 201412 I 39
SC 201412 J 18
SC 201412 K 6
SC 201501 B 5
SC 201501 C 60
SC 201501 D 123
SC 201501 E 114
SC 201501 F 96
SC 201501 G 79
SC 201501 H 78
SC 201501 I 58
SC 201501 J 21
SC 201501 K 6
SC 201502 B 11
SC 201502 C 60
SC 201502 D 105
SC 201502 E 73
SC 201502 F 72
SC 201502 G 69
SC 201502 H 74
SC 201502 I 47
SC 201502 J 17
SC 201502 K 6
SC 201503 B 6
SC 201503 C 72
SC 201503 D 114
SC 201503 E 96
SC 201503 F 52
SC 201503 G 69
SC 201503 H 86
SC 201503 I 47
SC 201503 J 17
SC 201503 K 7
SC 201504 B 11
SC 201504 C 65
SC 201504 D 127
SC 201504 E 85
SC 201504 F 92
SC 201504 G 71
SC 201504 H 77
SC 201504 I 38
SC 201504 J 16
SC 201504 K 3
SC 201505 B 20
SC 201505 C 87
SC 201505 D 142
SC 201505 E 124
SC 201505 F 89
SC 201505 G 87
SC 201505 H 92
SC 201505 I 57
SC 201505 J 28
SC 201505 K 13
SC 201506 B 22
SC 201506 C 69
SC 201506 D 117
SC 201506 E 97
SC 201506 F 92
SC 201506 G 75
SC 201506 H 64
SC 201506 I 49
SC 201506 J 12
SC 201506 K 9
SC 201507 B 20
SC 201507 C 97
SC 201507 D 141
SC 201507 E 119
SC 201507 F 110
SC 201507 G 96
SC 201507 H 82
SC 201507 I 37
SC 201507 J 21
SC 201507 K 7
SC 201508 B 25
SC 201508 C 100
SC 201508 D 135
SC 201508 E 112
SC 201508 F 92
SC 201508 G 93
SC 201508 H 74
SC 201508 I 44
SC 201508 J 15
SC 201508 K 5
SC 201509 B 11
SC 201509 C 65
SC 201509 D 128
SC 201509 E 117
SC 201509 F 100
SC 201509 G 94
SC 201509 H 73
SC 201509 I 48
SC 201509 J 16
SC 201509 K 9

Review Visualizations post-Data removal and second round of sampling¶

%%time


cols = list(SampledOPMDataProf.select_dtypes(include=['float64', 'int64']))
cols.remove('BLS_FEDERAL_OtherSep_Rate')
cols.remove('BLS_FEDERAL_Quits_Rate')
cols.remove('BLS_FEDERAL_TotalSep_Level')
cols.remove('BLS_FEDERAL_JobOpenings_Rate')
cols.remove('BLS_FEDERAL_OtherSep_Level')
cols.remove('BLS_FEDERAL_Quits_Level')
cols.remove('BLS_FEDERAL_JobOpenings_Level')
cols.remove('BLS_FEDERAL_Layoffs_Rate')
cols.remove('BLS_FEDERAL_Layoffs_Level')
cols.remove('BLS_FEDERAL_TotalSep_Rate')
cols.append('SEP')
display(cols)

plotNumeric = SampledOPMDataProf[cols]

# Create binary separation attribute for EDA correlation review
#plotNumeric["SEP_bin"] = plotNumeric.SEP.replace("NS", 1)
#plotNumeric.loc[plotNumeric['SEP_bin'] != 1, 'SEP_bin'] = 0
#plotNumeric.SEP_bin = plotNumeric.SEP_bin.apply(pd.to_numeric)
AttSplit = pd.get_dummies(plotNumeric['SEP'],prefix='SEP')
display(AttSplit.head())
plotNumeric = pd.concat((plotNumeric,AttSplit),axis=1) # add back into the dataframe

display(plotNumeric.head())
print("plotNumeric has {0} Records".format(len(plotNumeric)))
#print(plotNumeric.SEP_bin.dtype)

['GSEGRD',
 'SALARY',
 'LOS',
 'SEPCount_EFDATE_OCC',
 'SEPCount_EFDATE_LOC',
 'IndAvgSalary',
 'SalaryOverUnderIndAvg',
 'LowerLimitAge',
 'YearsToRetirement',
 'SALARYLog',
 'LOSSqrt',
 'SEPCount_EFDATE_OCCLog',
 'SEPCount_EFDATE_LOCLog',
 'IndAvgSalaryLog',
 'SEP']

plotNumeric has 14920 Records
CPU times: user 37.1 ms, sys: 1.19 ms, total: 38.3 ms
Wall time: 36.3 ms

%%time

sns.set(font_scale=1)
sns.pairplot(plotNumeric.drop(['SEP_NS',
                               'SEP_SC'], axis=1), hue = 'SEP', palette="hls", plot_kws={"s": 50})

/usr/local/es7/lib/python3.5/site-packages/matplotlib/font_manager.py:1297: UserWarning: findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans
  (prop.get_family(), self.defaultFamily[fontext]))

CPU times: user 45.7 s, sys: 41.6 s, total: 1min 27s
Wall time: 37.1 s

%%time

# Function modified from https://stackoverflow.com/questions/29530355/plotting-multiple-histograms-in-grid
sns.set()

def draw_histograms(df, variables, n_rows, n_cols):
    fig=plt.figure(figsize=(20,20))
    for i, var_name in enumerate(variables):
        ax=fig.add_subplot(n_rows,n_cols,i+1)
        df[var_name].hist(bins=20,ax=ax, color='#58D68D')
        ax.set_title(var_name+" Distribution")
    fig.tight_layout()  # Improves appearance a bit.
    plt.show()

draw_histograms(plotNumeric.drop(['SEP',
                                  'SEP_NS',
                                  'SEP_SC'
                                 ], axis=1),
                plotNumeric.drop(['SEP',
                                  'SEP_NS',
                                  'SEP_SC'
                                 ], axis=1).columns, 6, 3)

/usr/local/es7/lib/python3.5/site-packages/matplotlib/font_manager.py:1297: UserWarning: findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans
  (prop.get_family(), self.defaultFamily[fontext]))

CPU times: user 4.15 s, sys: 2.77 s, total: 6.91 s
Wall time: 3.64 s

%%time
# Inspired by http://seaborn.pydata.org/examples/many_pairwise_correlations.html

#plt.matshow(plotNumeric.corr())

sns.set(style='white')
corr = plotNumeric.drop(['SEP'], axis=1).corr()

# Generate a mask for the upper triangle
mask = np.zeros_like(corr, dtype=np.bool)
mask[np.triu_indices_from(mask, k=1)] = True

# Set up the matplotlib figure
f, ax = plt.subplots(figsize=(20, 20))

# Generate a custom diverging colormap
cmap = sns.diverging_palette(250, 10, as_cmap=True)

# Draw the heatmap with the mask and correct aspect ratio
sns.set(font_scale=0.95)
heatCorr = sns.heatmap(corr, mask=mask, cmap=cmap, vmax=1, vmin=-1,
                       square=True, annot=True, linewidths=1,
                       cbar_kws={"shrink": .5}, ax=ax, fmt='.1g')
#heatCorr.
ax.tick_params(labelsize=15)
cax = plt.gcf().axes[-1]
cax.tick_params(labelsize=15)

sns.plt.show()
#sns.heatmap(corr, annot=True, linewidths=0.01, cmap=cmap, ax=ax)

/usr/local/es7/lib/python3.5/site-packages/matplotlib/font_manager.py:1297: UserWarning: findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans
  (prop.get_family(), self.defaultFamily[fontext]))

CPU times: user 1.64 s, sys: 870 ms, total: 2.51 s
Wall time: 1.49 s

%%time

cols = list(SampledOPMDataProf.select_dtypes(include=['object']))
dropCols = ["LOCTYP",
            "LOCTYPT",
            "OCCTYP",
            "OCCTYPT",
            "PPTYP",
            "PPTYPT",
            "AGYTYP",
            "OCCFAM",
            "PPGROUP",
            "PAYPLAN",
            "TOATYP",
            "WSTYP",
            "AGYSUBT",
            "AGELVL",
            "LOSLVL",
            "LOC",
            "OCC",
            "PATCO",
            "SALLVL",
            "TOA",
            "WORKSCH"]

for i in dropCols:
    if(i in list(SampledOPMDataProf.columns)): cols.remove(i)

plotCat = SampledOPMDataProf[cols]
display(plotCat.head())
print("plotCat Has {0} Records".format(len(plotCat)))
print("Number of colums = ", len(cols))

plotCat Has 14920 Records
Number of colums =  11
CPU times: user 15.3 ms, sys: 2.8 ms, total: 18.1 ms
Wall time: 16.2 ms

%%time

for i in cols:
    if i != 'SEP':
        plt.figure(i) # Required to create new figure each loop rather than drawing over previous object
        f, (ax1, ax2) = plt.subplots(ncols=2, figsize=(20, 10), sharey=False)
        sns.countplot(y=i, data=plotCat, color="lightblue", ax=ax1);
        sns.countplot(y=i, data=plotCat, hue="SEP", palette="hls", ax=ax2);
        
    if i == 'AGYSUB':
        subCountPlot(i, 'SEP', 10000)
    elif i == 'LOCT':
        subCountPlot(i, 'SEP', 1000)
    elif i == 'OCCT':
        subCountPlot(i, 'SEP', 2000)
    elif i == 'PPGRD':
        subCountPlot(i, 'SEP', 6000)
    elif i == 'AGYT':
        subCountPlot(i, 'SEP', 3000)

CPU times: user 2.1 s, sys: 28.4 ms, total: 2.13 s
Wall time: 2.11 s

/usr/local/es7/lib/python3.5/site-packages/matplotlib/pyplot.py:524: RuntimeWarning: More than 20 figures have been opened. Figures created through the pyplot interface (`matplotlib.pyplot.figure`) are retained until explicitly closed and may consume too much memory. (To control this warning, see the rcParam `figure.max_open_warning`).
  max_open_warning, RuntimeWarning)

<matplotlib.figure.Figure at 0x7f6ca3817198>

/usr/local/es7/lib/python3.5/site-packages/matplotlib/font_manager.py:1297: UserWarning: findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans
  (prop.get_family(), self.defaultFamily[fontext]))

<matplotlib.figure.Figure at 0x7f71aabdc470>

<matplotlib.figure.Figure at 0x7f71b60eda20>

<matplotlib.figure.Figure at 0x7f728565e080>

<matplotlib.figure.Figure at 0x7f71b0f84518>

<matplotlib.figure.Figure at 0x7f72858010b8>

<matplotlib.figure.Figure at 0x7f7198c1deb8>

%%time

for i in cols:
    if i != 'SEP':
        percBarPlot(i, 'SEP', len(plotCat.SEP.drop_duplicates()))

CPU times: user 1.08 s, sys: 45.4 ms, total: 1.12 s
Wall time: 1.06 s

/usr/local/es7/lib/python3.5/site-packages/matplotlib/font_manager.py:1297: UserWarning: findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans
  (prop.get_family(), self.defaultFamily[fontext]))

%%time

sns.set(style="whitegrid", palette="pastel", color_codes=True)

sns.violinplot(x="PATCOT", y="SALARY", data=SampledOPMDataProf, split=True,
               inner="quart")
sns.despine(left=True)

CPU times: user 1.66 s, sys: 8.67 s, total: 10.3 s
Wall time: 342 ms

/usr/local/es7/lib/python3.5/site-packages/matplotlib/font_manager.py:1297: UserWarning: findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans
  (prop.get_family(), self.defaultFamily[fontext]))

%%time

# Draw a nested violinplot and split the violins for easier comparison
sns.violinplot(x="SEP", y="SALARY", data=SampledOPMDataProf, split=True,
               inner="box", scale="area", cut=0)
sns.despine(left=True)

CPU times: user 160 ms, sys: 111 ms, total: 271 ms
Wall time: 135 ms

/usr/local/es7/lib/python3.5/site-packages/matplotlib/font_manager.py:1297: UserWarning: findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans
  (prop.get_family(), self.defaultFamily[fontext]))

#%%time
#
#sns.factorplot(x="SEP", y="SALARY", col="PATCOT",
#               data=SampledOPMDataProf,
#               kind="violin", split=True, aspect=.5, size=15);

#%%time
#
#sns.factorplot(x="SEP", y="SALARY", col="PATCOT", data=SampledOPMDataProf,
#               kind="violin", split=True, aspect=.4, size=10);

%%time

g = sns.PairGrid(data=SampledOPMDataProf,
                 x_vars=["SEP","PATCOT"],
                 y_vars=["SALARY", "LOS", "LowerLimitAge", "YearsToRetirement"],
                 aspect=1, size=10)
g.map(sns.violinplot, palette="pastel", inner="quart");

/usr/local/es7/lib/python3.5/site-packages/matplotlib/font_manager.py:1297: UserWarning: findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans
  (prop.get_family(), self.defaultFamily[fontext]))

CPU times: user 4.68 s, sys: 23.7 s, total: 28.4 s
Wall time: 1.35 s

Encode Categorical Attributes, and Remove Description Columns for Analysis Prep¶

Now that we have the dataset sampled, we still have some legwork necessary to convert our categorical attributes into binary integer values. Below we walk through this process for the following Attributes:

AGELVL
LOC
SALLVL
TOA
OCCTYP
OCCFAM
PPTYP
PPGROUP
TOATYP

Once these attributes have been encoded and description columns removed, we end up with a total of 2446 attributes in our dataset for analysis in our model generation.

# Clean up old objects no longer needed, to clear up memory
process = psutil.Process(os.getpid())
print("Memory Usage before Cleanup: ", process.memory_info().rss)

if 'AGELVL' in dir():
    del AGELVL
if 'AggIndAvgSalary' in dir():
    del AggIndAvgSalary
if 'AggIndAvgSalary2' in dir():
    del AggIndAvgSalary2
if 'AggSEPCount_EFDATE_LOC' in dir():
    del AggSEPCount_EFDATE_LOC
if 'AggSEPCount_EFDATE_OCC' in dir():
    del AggSEPCount_EFDATE_OCC
if 'AggStrat' in dir():
    del AggStrat
if 'DATECODE' in dir():
    del DATECODE
if 'EMPColList' in dir():
    del EMPColList
if 'EMPDataOrig4Q' in dir():
    del EMPDataOrig4Q
if 'maxSize' in dir():
    del maxSize
if 'OPMColList' in dir():
    del OPMColList
if 'OPMDataFiles' in dir():
    del OPMDataFiles
if 'OPMDataList' in dir():
    del OPMDataList
if 'OPMDataMerged' in dir():
    del OPMDataMerged
if 'OPMDataOrig' in dir():
    del OPMDataOrig
if 'SEP' in dir():
    del SEP
if 'SampleSize' in dir():
    del SampleSize
if 'SampledOPMStratumData' in dir():
    del SampledOPMStratumData
if 'SampledOPMStratumDataList' in dir():
    del SampledOPMStratumDataList
if 'StratCountSample' in dir():
    del StratCountSample
if 'StratSampleSize' in dir():
    del StratSampleSize
if 'JTL' in dir():
    del JTL
    
process = psutil.Process(os.getpid())
print("Memory Usage after Cleanup: ", process.memory_info().rss)

Memory Usage before Cleanup:  14041321472
Memory Usage after Cleanup:  12331384832

display(SampledOPMDataProf.head())
SampledOPMDataProf.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 14920 entries, 0 to 14919
Data columns (total 50 columns):
SEP                              14920 non-null object
DATECODE                         14920 non-null object
AGELVL                           14920 non-null object
GSEGRD                           14920 non-null float64
LOC                              14920 non-null object
PATCO                            14920 non-null object
TOA                              14920 non-null object
WORKSCH                          14920 non-null object
SALARY                           14920 non-null float64
LOS                              14920 non-null float64
AGELVLT                          14920 non-null object
LOCTYP                           14920 non-null object
LOCTYPT                          14920 non-null object
LOCT                             14920 non-null object
OCCTYP                           14920 non-null object
OCCTYPT                          14920 non-null object
OCCFAM                           14920 non-null object
OCCFAMT                          14920 non-null object
PATCOT                           14920 non-null object
PPTYP                            14920 non-null object
PPTYPT                           14920 non-null object
PPGROUP                          14920 non-null object
PPGROUPT                         14920 non-null object
TOATYP                           14920 non-null object
TOATYPT                          14920 non-null object
TOAT                             14920 non-null object
WSTYP                            14920 non-null object
WSTYPT                           14920 non-null object
WORKSCHT                         14920 non-null object
SEPCount_EFDATE_OCC              14920 non-null float64
SEPCount_EFDATE_LOC              14920 non-null int64
IndAvgSalary                     14920 non-null float64
SalaryOverUnderIndAvg            14920 non-null float64
LowerLimitAge                    14920 non-null float64
YearsToRetirement                14920 non-null float64
BLS_FEDERAL_OtherSep_Rate        14920 non-null float64
BLS_FEDERAL_Quits_Rate           14920 non-null float64
BLS_FEDERAL_TotalSep_Level       14920 non-null int64
BLS_FEDERAL_JobOpenings_Rate     14920 non-null float64
BLS_FEDERAL_OtherSep_Level       14920 non-null int64
BLS_FEDERAL_Quits_Level          14920 non-null int64
BLS_FEDERAL_JobOpenings_Level    14920 non-null int64
BLS_FEDERAL_Layoffs_Rate         14920 non-null float64
BLS_FEDERAL_Layoffs_Level        14920 non-null int64
BLS_FEDERAL_TotalSep_Rate        14920 non-null float64
SALARYLog                        14920 non-null float64
LOSSqrt                          14920 non-null float64
SEPCount_EFDATE_OCCLog           14920 non-null float64
SEPCount_EFDATE_LOCLog           14920 non-null float64
IndAvgSalaryLog                  14920 non-null float64
dtypes: float64(18), int64(6), object(26)
memory usage: 5.7+ MB

%%time

if os.path.isfile(PickleJarPath+"/OPMAnalysisDataNoFamBinary.pkl"):
    print("Found the File! Loading Pickle Now!")
    OPMAnalysisDataNoFamBinary = unpickleObject("OPMAnalysisDataNoFamBinary")
else:

    OPMAnalysisDataNoFamBinary = SampledOPMDataProf.copy()

    cols = ["GENDER",
            "DATECODE",
            "QTR",
            "COUNT",
            "AGYTYPT",
            "AGYT",
            "AGYSUB",
            "AGYSUBT",
            "QTR",
            "AGELVLT",
            "LOSLVL",
            "LOSLVLT",
            "LOCTYPT",
            "LOCT",
            "OCCTYP",
            "OCCTYPT",
            "OCCFAM",
            "OCCFAMT",
            "OCC",
            "OCCT",
            "PATCO",
            "PPGRD",
            "PATCOT",
            "PPTYPT",
            "PPGROUPT",
            "PAYPLAN",
            "PAYPLANT",
            "SALLVLT",
            "TOATYPT",
            "TOAT",
            "WSTYP",
            "WSTYPT",
            "WORKSCH",
            "WORKSCHT",
            "SALARY",
            "LOS",
            "SEPCount_EFDATE_OCC",
            "SEPCount_EFDATE_LOC"
           ]



    #delete cols from analysis data
    for col in cols:
        if col in list(OPMAnalysisDataNoFamBinary.columns):
            del OPMAnalysisDataNoFamBinary[col]

    OPMAnalysisDataNoFamBinary.info()

    cols = ["AGELVL",
            "LOC",
            "SALLVL",
            "TOA",
            "AGYTYP",
            "AGY",
            "LOCTYP",
            "PPTYP",
            "PPGROUP",
            "TOATYP"
           ]

    #Split Values for cols 
    for col in cols:
        if col in list(OPMAnalysisDataNoFamBinary.columns):
            AttSplit = pd.get_dummies(OPMAnalysisDataNoFamBinary[col],prefix=col)
            display(AttSplit.head())
            OPMAnalysisDataNoFamBinary = pd.concat((OPMAnalysisDataNoFamBinary,AttSplit),axis=1) # add back into the dataframe
            del OPMAnalysisDataNoFamBinary[col]

    pickleObject(OPMAnalysisDataNoFamBinary, "OPMAnalysisDataNoFamBinary")
        
display(OPMAnalysisDataNoFamBinary.head())
print("Number of Columns: ",len(OPMAnalysisDataNoFamBinary.columns))
OPMAnalysisDataNoFamBinary.info()

Found the File! Loading Pickle Now!

Number of Columns:  100
<class 'pandas.core.frame.DataFrame'>
RangeIndex: 14920 entries, 0 to 14919
Data columns (total 100 columns):
SEP                              14920 non-null object
GSEGRD                           14920 non-null float64
IndAvgSalary                     14920 non-null float64
SalaryOverUnderIndAvg            14920 non-null float64
LowerLimitAge                    14920 non-null float64
YearsToRetirement                14920 non-null float64
BLS_FEDERAL_OtherSep_Rate        14920 non-null float64
BLS_FEDERAL_Quits_Rate           14920 non-null float64
BLS_FEDERAL_TotalSep_Level       14920 non-null int64
BLS_FEDERAL_JobOpenings_Rate     14920 non-null float64
BLS_FEDERAL_OtherSep_Level       14920 non-null int64
BLS_FEDERAL_Quits_Level          14920 non-null int64
BLS_FEDERAL_JobOpenings_Level    14920 non-null int64
BLS_FEDERAL_Layoffs_Rate         14920 non-null float64
BLS_FEDERAL_Layoffs_Level        14920 non-null int64
BLS_FEDERAL_TotalSep_Rate        14920 non-null float64
SALARYLog                        14920 non-null float64
LOSSqrt                          14920 non-null float64
SEPCount_EFDATE_OCCLog           14920 non-null float64
SEPCount_EFDATE_LOCLog           14920 non-null float64
IndAvgSalaryLog                  14920 non-null float64
AGELVL_B                         14920 non-null uint8
AGELVL_C                         14920 non-null uint8
AGELVL_D                         14920 non-null uint8
AGELVL_E                         14920 non-null uint8
AGELVL_F                         14920 non-null uint8
AGELVL_G                         14920 non-null uint8
AGELVL_H                         14920 non-null uint8
AGELVL_I                         14920 non-null uint8
AGELVL_J                         14920 non-null uint8
AGELVL_K                         14920 non-null uint8
LOC_01                           14920 non-null uint8
LOC_02                           14920 non-null uint8
LOC_04                           14920 non-null uint8
LOC_05                           14920 non-null uint8
LOC_06                           14920 non-null uint8
LOC_08                           14920 non-null uint8
LOC_09                           14920 non-null uint8
LOC_10                           14920 non-null uint8
LOC_11                           14920 non-null uint8
LOC_12                           14920 non-null uint8
LOC_13                           14920 non-null uint8
LOC_15                           14920 non-null uint8
LOC_16                           14920 non-null uint8
LOC_17                           14920 non-null uint8
LOC_18                           14920 non-null uint8
LOC_19                           14920 non-null uint8
LOC_20                           14920 non-null uint8
LOC_21                           14920 non-null uint8
LOC_22                           14920 non-null uint8
LOC_23                           14920 non-null uint8
LOC_24                           14920 non-null uint8
LOC_25                           14920 non-null uint8
LOC_26                           14920 non-null uint8
LOC_27                           14920 non-null uint8
LOC_28                           14920 non-null uint8
LOC_29                           14920 non-null uint8
LOC_30                           14920 non-null uint8
LOC_31                           14920 non-null uint8
LOC_32                           14920 non-null uint8
LOC_33                           14920 non-null uint8
LOC_34                           14920 non-null uint8
LOC_35                           14920 non-null uint8
LOC_36                           14920 non-null uint8
LOC_37                           14920 non-null uint8
LOC_38                           14920 non-null uint8
LOC_39                           14920 non-null uint8
LOC_40                           14920 non-null uint8
LOC_41                           14920 non-null uint8
LOC_42                           14920 non-null uint8
LOC_44                           14920 non-null uint8
LOC_45                           14920 non-null uint8
LOC_46                           14920 non-null uint8
LOC_47                           14920 non-null uint8
LOC_48                           14920 non-null uint8
LOC_49                           14920 non-null uint8
LOC_50                           14920 non-null uint8
LOC_51                           14920 non-null uint8
LOC_53                           14920 non-null uint8
LOC_54                           14920 non-null uint8
LOC_55                           14920 non-null uint8
LOC_56                           14920 non-null uint8
TOA_10                           14920 non-null uint8
TOA_15                           14920 non-null uint8
TOA_20                           14920 non-null uint8
TOA_30                           14920 non-null uint8
TOA_32                           14920 non-null uint8
TOA_35                           14920 non-null uint8
TOA_38                           14920 non-null uint8
TOA_40                           14920 non-null uint8
TOA_42                           14920 non-null uint8
TOA_44                           14920 non-null uint8
TOA_45                           14920 non-null uint8
TOA_48                           14920 non-null uint8
LOCTYP_1                         14920 non-null uint8
PPTYP_1                          14920 non-null uint8
PPGROUP_11                       14920 non-null uint8
PPGROUP_12                       14920 non-null uint8
TOATYP_1                         14920 non-null uint8
TOATYP_2                         14920 non-null uint8
dtypes: float64(15), int64(5), object(1), uint8(79)
memory usage: 3.5+ MB
CPU times: user 92.1 ms, sys: 2.09 ms, total: 94.2 ms
Wall time: 91.9 ms

Below is a display of all remaining attributes and their corresponding data types for analysis

%%time

data_type = []
for idx, col in enumerate(OPMAnalysisDataNoFamBinary.columns):
    data_type.append(OPMAnalysisDataNoFamBinary.dtypes[idx])

summary_df = {'Attribute Name' : pd.Series(OPMAnalysisDataNoFamBinary.columns, index = range(len(OPMAnalysisDataNoFamBinary.columns))), 'Data Type' : pd.Series(data_type, index = range(len(OPMAnalysisDataNoFamBinary.columns)))}
summary_df = pd.DataFrame(summary_df)
display(summary_df)

del data_type, summary_df

CPU times: user 24.3 ms, sys: 89 µs, total: 24.4 ms
Wall time: 23.7 ms

Dimensionality Reduction using Principal Component Analysis¶

We also scale the data values to remove bias in our models due to different attribute scales. Without scaling the data, attributes such as SALARY and LOS would carry heavier weights when compared against the binary encoded attributes and BLS data. This would cause unbalanced and improperly analyzed data for model creation.

OPMScaledAnalysisData = OPMAnalysisDataNoFamBinary.copy()
del OPMScaledAnalysisData["SEP"]

%%time

OPMAnalysisScalerFit = MinMaxScaler().fit(OPMScaledAnalysisData)
## Pickle for later re-use if needed
pickleObject(OPMAnalysisScalerFit, "OPMAnalysisScalerFit")

OPMScaledAnalysisData = pd.DataFrame(OPMAnalysisScalerFit.transform(OPMScaledAnalysisData), columns = OPMScaledAnalysisData.columns)

CPU times: user 12.5 ms, sys: 834 µs, total: 13.4 ms
Wall time: 14.4 ms

display(OPMScaledAnalysisData.head())

PCA Principal Components defined¶

Our objective, is to reduce dimensionality through identification of principal components. We have chosen to use the full column input (99) as the maximum number of components to be produced. Given our hopes are to reduce the number of attributes needed for a model, we expect to find much smaller than 99 as our Principal components which explain over 80% variance within the dataset. We will review each component's explained variance further to determine the proper number of components to be included later during model generation. Note randomized PCA was chosen in order to use singular value decomposition in our dimensionality reduction efforts due to the large size of our data set.

%%time

seed = len(OPMScaledAnalysisData)

print(OPMScaledAnalysisData.shape)
pca_class = PCA(n_components=len(OPMScaledAnalysisData.columns), svd_solver='randomized', random_state=seed)

pca_class.fit(OPMScaledAnalysisData)

(14920, 99)
CPU times: user 11.4 s, sys: 16.6 s, total: 28 s
Wall time: 635 ms

Below, the resulting components have been ordered by eigenvector value and these values portrayed as ratios of variance explained by each component. In order to identify the principal components to be included during model generation, we review the rate at which explained variance decreases in significance from one principal component to the next. Accompanying these proportion values is a scree plot representing these same values in visual form. By plotting the scree plot, it is easier to judge where this rate of decreasing explained variance occurs. Note the rate of change in explained variance among the first 8 principal components, with another less significant change through the 22th component. After the 22th component, the rate of decreasing explained variance begins to somewhat flatten out.

%%time

#The amount of variance that each PC explains
var= pca_class.explained_variance_ratio_

sns.set(font_scale=1.7)
plt.plot(range(1,len(OPMScaledAnalysisData.columns)+1), var*100, marker = '.', color = 'red', markerfacecolor = 'black')
plt.xlabel('Principal Components')
plt.ylabel('Percentage of Explained Variance')
plt.title('Scree Plot')
plt.axis([0, len(OPMScaledAnalysisData.columns)+1, -0.1, 9])
plt.annotate('22nd Component', xy=(22, 1.2), xytext=(40, 4),
                arrowprops=dict(facecolor='black', shrink=0.05),)
np.set_printoptions(suppress=True)
print(np.round(var, decimals=4)*100)

[ 11.45   9.85   6.05   5.42   4.49   3.84   3.61   3.41   3.13   2.99
   2.76   2.74   2.72   2.6    2.53   2.36   2.31   2.05   1.89   1.68
   1.64   1.18   1.09   0.96   0.93   0.91   0.81   0.78   0.68   0.66
   0.64   0.61   0.58   0.54   0.51   0.49   0.46   0.45   0.44   0.41
   0.39   0.39   0.36   0.36   0.3    0.3    0.29   0.28   0.28   0.28
   0.26   0.22   0.21   0.21   0.21   0.21   0.2    0.2    0.2    0.19
   0.18   0.17   0.17   0.16   0.13   0.13   0.11   0.11   0.1    0.1    0.1
   0.1    0.07   0.07   0.06   0.06   0.04   0.04   0.03   0.02   0.02
   0.02   0.01   0.01   0.     0.     0.     0.     0.     0.     0.     0.
   0.     0.     0.     0.     0.     0.     0.  ]
CPU times: user 359 ms, sys: 2.27 s, total: 2.63 s
Wall time: 68 ms

/usr/local/es7/lib/python3.5/site-packages/matplotlib/font_manager.py:1297: UserWarning: findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans
  (prop.get_family(), self.defaultFamily[fontext]))

By now referring to the cumulative variance values and associated plot below, it may be seen that the cumulative variance increases in a fairly consistent parabola curve. In attempts to acheive a cumulative variance explained of greater than 80%, we end at 22 principal components. For this reason, 22 principal components may be selected as being the most appropriate for separation classification modeling given the variables among these data.

#Cumulative Variance explains
var1=np.cumsum(np.round(pca_class.explained_variance_ratio_, decimals=4)*100)

plt.plot(range(1,len(OPMScaledAnalysisData.columns)+1), var1, marker = '.', color = 'green', markerfacecolor = 'black')
plt.xlabel('Principal Components')
plt.ylabel('Explained Variance (Sum %)')
plt.title('Cumulative Variance Plot')
plt.axis([0, len(OPMScaledAnalysisData.columns)+1, 10, len(OPMScaledAnalysisData.columns)+1])
plt.annotate('22nd Component', xy=(22, 80.54), xytext=(40, 60),
                arrowprops=dict(facecolor='black', shrink=0.05),)
print(var1)

[ 11.45  21.3   27.35  32.77  37.26  41.1   44.71  48.12  51.25  54.24  57.
  59.74  62.46  65.06  67.59  69.95  72.26  74.31  76.2   77.88  79.52
  80.7   81.79  82.75  83.68  84.59  85.4   86.18  86.86  87.52  88.16
  88.77  89.35  89.89  90.4   90.89  91.35  91.8   92.24  92.65  93.04
  93.43  93.79  94.15  94.45  94.75  95.04  95.32  95.6   95.88  96.14
  96.36  96.57  96.78  96.99  97.2   97.4   97.6   97.8   97.99  98.17
  98.34  98.51  98.67  98.8   98.93  99.04  99.15  99.25  99.35  99.45
  99.55  99.62  99.69  99.75  99.81  99.85  99.89  99.92  99.94  99.96
  99.98  99.99 100.   100.   100.   100.   100.   100.   100.   100.   100.
 100.   100.   100.   100.   100.   100.   100.  ]

/usr/local/es7/lib/python3.5/site-packages/matplotlib/font_manager.py:1297: UserWarning: findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans
  (prop.get_family(), self.defaultFamily[fontext]))

We proceed to analyze the first 4 component Feature Loadings more carefully. See below, plots of the top 10 loadings for each component.

plt.rcParams['figure.figsize'] = (20, 12)
fig = plt.figure()
plt.rcParams.update({'font.size': 16})
plt.rc('xtick', labelsize=15)
plt.rc('ytick', labelsize=15) 
for i in range(0,4):
    components = pd.Series(pca_class.components_[i], index=OPMScaledAnalysisData.columns)

    maxcomponent = pd.Series(pd.DataFrame(abs(components).sort_values(ascending=False).head(10)).index)

    matplotlib.rc('xtick', labelsize=12)


    ax = fig.add_subplot(2,2,i + 1)
       
    weightsplot = pd.Series(components, index=maxcomponent)
    weightsplot.plot(title = "Principal Component "+ str(i+1), kind='bar', color = 'Tomato', ax = ax)

plt.tight_layout()
plt.show()

/usr/local/es7/lib/python3.5/site-packages/matplotlib/font_manager.py:1297: UserWarning: findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans
  (prop.get_family(), self.defaultFamily[fontext]))

MaxPC = 22

PCList = []
for i in range(0,MaxPC):
    components = pd.Series(pca_class.components_[i], index=OPMScaledAnalysisData.columns)

    maxcomponent = pd.Series(pd.DataFrame(abs(components).sort_values(ascending=False).head(15)).index)

    PCList.append(maxcomponent)

PCList = pd.concat(PCList).drop_duplicates().sort_values(ascending=True).reset_index(drop = True)
print(PCList)
PCList = list(PCList)

0                          AGELVL_C
1                          AGELVL_D
2                          AGELVL_E
3                          AGELVL_F
4                          AGELVL_G
5                          AGELVL_H
6                          AGELVL_I
7                          AGELVL_J
8                          AGELVL_K
9     BLS_FEDERAL_JobOpenings_Level
10     BLS_FEDERAL_JobOpenings_Rate
11        BLS_FEDERAL_Layoffs_Level
12         BLS_FEDERAL_Layoffs_Rate
13       BLS_FEDERAL_OtherSep_Level
14        BLS_FEDERAL_OtherSep_Rate
15          BLS_FEDERAL_Quits_Level
16           BLS_FEDERAL_Quits_Rate
17       BLS_FEDERAL_TotalSep_Level
18        BLS_FEDERAL_TotalSep_Rate
19                           GSEGRD
20                     IndAvgSalary
21                  IndAvgSalaryLog
22                           LOC_04
23                           LOC_06
24                           LOC_08
25                           LOC_11
26                           LOC_12
27                           LOC_13
28                           LOC_24
29                           LOC_36
30                           LOC_48
31                           LOC_51
32                          LOSSqrt
33                    LowerLimitAge
34                       PPGROUP_11
35                       PPGROUP_12
36                        SALARYLog
37           SEPCount_EFDATE_LOCLog
38           SEPCount_EFDATE_OCCLog
39                         TOATYP_1
40                         TOATYP_2
41                           TOA_10
42                           TOA_15
43                           TOA_20
44                           TOA_30
45                           TOA_38
46                           TOA_40
47                           TOA_48
48                YearsToRetirement
dtype: object

Total of 50 features of the original 99 are identified, by taking the top 15 feature loadings within the first 22 components as determined above as the appropriate components to maximize variance explained. We may now, optionally utilize these 50 features identified, or utilize principal component vectors for analysis in the next steps.

Separation Response Weights¶

Due to the unproportional number of observations in each separation type in our dataset, we need to create weightings. using SciKit's class_weight algorithm, we compute an array of weights to be used downstream in our models.

Predicting Separation¶

We have chosen to utilize Stratified KFold Cross Validation for our classification analysis, with 5 folds. This means, that from our original sample size of 16,638, each "fold" will save off approximately 20% as test observations utilizing the rest as training observations all while keeping the ratio of classes equal amongst customers and subscribers. This process will occur through 5 iterations, or folds, to allow us to cross validate our results amongst different test/train combinations. We have utilized a random_state seed equal to the length of the original sampled dataset to ensure reproducible results.

seed = len(OPMAnalysisDataNoFamBinary)

cv = StratifiedKFold(n_splits = 5, random_state = seed)
print(OPMAnalysisDataNoFamBinary.shape)
print(cv)

(14920, 100)
StratifiedKFold(n_splits=5, random_state=14920, shuffle=False)

Random Forest Classification¶

Max Depth The maximum depth (levels) in the tree. When a value is set, the tree may not split further once this level has been met regardless of how many nodes are in the leaf.

Max Features Number of features to consider when looking for a split.

Minimum Samples in Leaf Minimum number of samples required to be in a leaf node. Splits may not occur which cause the number of samples in a leaf to be less than this value. Too low a value here leads to overfitting the tree to train data.

Minimum Samples to Split Minimum number fo samples required to split a node. Care was taken during parameter tests to keep the ratio between Min Samples in Leaf and Min Samples to Split equal to that of the default values (1:2). This was done to allow an even 50/50 split on nodes which match the lowest granularity split criteria. similar to the min samples in leaf, too low a value here leads to overfitting the tree to train data.

n_estimators Number of Trees generated in the forest. Increasing the number of trees, in our models increased accuracy while decreasing performance. We tuned to provide output that completed all 10 iterations in under 10 minutes.

Not Complete#### After 13 iterations of modifying the above parameters, we land on a final winner based on the highest average Accuracy value across all iterations. Average Accuracy values in our 10 test/train iterations ranged from 70.2668 % from default inputs of the random forest classification model to a value of 72.5192 % in the best tuned model fit. Although the run-time of this model parameter choice is the largest performed, we decided to remain with these inputs due to the amount increase in accuracy. As mentioned previously, we tuned the n_estimators parameter to ensure we stayed under 10 minutes execution. Parameter inputs for the final Random Forest Classification model with the KD Tree Algorithm are as follows: ###Not Complete¶

max_depth	max_features	min_samples_leaf	min_samples_split	n_estimators
TBD	TBD	TBD	TBD	TBD

%%time
"""
def rfc_explor(n_estimators,
               max_features,
               max_depth, 
               min_samples_split,
               min_samples_leaf,
               Data        = OPMAnalysisDataNoFam,
               cols        = PCList,
               cv          = cv,
               seed        = seed):
    startTime = datetime.now()
    y = Data["SEP"].values # get the labels we want    
    
    X = Data[cols].as_matrix()
    
    rfc_clf = RandomForestClassifier(n_estimators=n_estimators, max_features = max_features, max_depth=max_depth, min_samples_split = min_samples_split, min_samples_leaf = min_samples_leaf, class_weight = "balanced", n_jobs=-1, random_state = seed) # get object
    
    # setup pipeline to take PCA, then fit a clf model
    clf_pipe = Pipeline(
        [('minMaxScaler', MinMaxScaler()),
         ('CLF',rfc_clf)]
    )

    accuracy = cross_val_score(clf_pipe, X, y, cv=cv.split(X, y)) # this also can help with parallelism
    MeanAccuracy =  sum(accuracy)/len(accuracy)
    accuracy = np.append(accuracy, MeanAccuracy)
    endTime = datetime.now()
    TotalTime = endTime - startTime
    accuracy = np.append(accuracy, TotalTime)
    
    print(TotalTime)
    print(accuracy)
    
    return accuracy
"""

CPU times: user 3 µs, sys: 1e+03 ns, total: 4 µs
Wall time: 8.58 µs

'\ndef rfc_explor(n_estimators,\n               max_features,\n               max_depth, \n               min_samples_split,\n               min_samples_leaf,\n               Data        = OPMAnalysisDataNoFam,\n               cols        = PCList,\n               cv          = cv,\n               seed        = seed):\n    startTime = datetime.now()\n    y = Data["SEP"].values # get the labels we want    \n    \n    X = Data[cols].as_matrix()\n    \n    rfc_clf = RandomForestClassifier(n_estimators=n_estimators, max_features = max_features, max_depth=max_depth, min_samples_split = min_samples_split, min_samples_leaf = min_samples_leaf, class_weight = "balanced", n_jobs=-1, random_state = seed) # get object\n    \n    # setup pipeline to take PCA, then fit a clf model\n    clf_pipe = Pipeline(\n        [(\'minMaxScaler\', MinMaxScaler()),\n         (\'CLF\',rfc_clf)]\n    )\n\n    accuracy = cross_val_score(clf_pipe, X, y, cv=cv.split(X, y)) # this also can help with parallelism\n    MeanAccuracy =  sum(accuracy)/len(accuracy)\n    accuracy = np.append(accuracy, MeanAccuracy)\n    endTime = datetime.now()\n    TotalTime = endTime - startTime\n    accuracy = np.append(accuracy, TotalTime)\n    \n    print(TotalTime)\n    print(accuracy)\n    \n    return accuracy\n'

%%time
"""
def rfc_explor_w_PCA(n_estimators,
               max_features,
               max_depth, 
               min_samples_split,
               min_samples_leaf,
               PCA,
               Data        = OPMAnalysisDataNoFam,
               cv          = cv,
               seed        = seed):
    startTime = datetime.now()
    y = Data["SEP"].values # get the labels we want    
    
    X = Data.drop("SEP", axis=1).as_matrix()
    
    rfc_clf = RandomForestClassifier(n_estimators=n_estimators, max_features = max_features, max_depth=max_depth, min_samples_split = min_samples_split, min_samples_leaf = min_samples_leaf, class_weight = "balanced", n_jobs=-1, random_state = seed) # get object
    
    # setup pipeline to take PCA, then fit a clf model
    clf_pipe = Pipeline(
        [('minMaxScaler', MinMaxScaler()),
         ('PCA', PCA),
         ('CLF',rfc_clf)]
    )

    accuracy = cross_val_score(clf_pipe, X, y, cv=cv.split(X, y)) # this also can help with parallelism
    MeanAccuracy =  sum(accuracy)/len(accuracy)
    accuracy = np.append(accuracy, MeanAccuracy)
    endTime = datetime.now()
    TotalTime = endTime - startTime
    accuracy = np.append(accuracy, TotalTime)
    
    #print(TotalTime)
    #print(accuracy)
    
    return accuracy
"""

CPU times: user 4 µs, sys: 1e+03 ns, total: 5 µs
Wall time: 8.11 µs

'\ndef rfc_explor_w_PCA(n_estimators,\n               max_features,\n               max_depth, \n               min_samples_split,\n               min_samples_leaf,\n               PCA,\n               Data        = OPMAnalysisDataNoFam,\n               cv          = cv,\n               seed        = seed):\n    startTime = datetime.now()\n    y = Data["SEP"].values # get the labels we want    \n    \n    X = Data.drop("SEP", axis=1).as_matrix()\n    \n    rfc_clf = RandomForestClassifier(n_estimators=n_estimators, max_features = max_features, max_depth=max_depth, min_samples_split = min_samples_split, min_samples_leaf = min_samples_leaf, class_weight = "balanced", n_jobs=-1, random_state = seed) # get object\n    \n    # setup pipeline to take PCA, then fit a clf model\n    clf_pipe = Pipeline(\n        [(\'minMaxScaler\', MinMaxScaler()),\n         (\'PCA\', PCA),\n         (\'CLF\',rfc_clf)]\n    )\n\n    accuracy = cross_val_score(clf_pipe, X, y, cv=cv.split(X, y)) # this also can help with parallelism\n    MeanAccuracy =  sum(accuracy)/len(accuracy)\n    accuracy = np.append(accuracy, MeanAccuracy)\n    endTime = datetime.now()\n    TotalTime = endTime - startTime\n    accuracy = np.append(accuracy, TotalTime)\n    \n    #print(TotalTime)\n    #print(accuracy)\n    \n    return accuracy\n'

%%time
"""
def rfc_explor_w_PCA(n_estimators,
               max_features,
               max_depth, 
               min_samples_split,
               min_samples_leaf,
               PCA,
               Data        = OPMAnalysisDataNoFam,
               cv          = cv,
               seed        = seed):
    startTime = datetime.now()
    y = Data["SEP"].values # get the labels we want    
    
    X = Data.drop("SEP", axis=1).as_matrix()
    
    rfc_clf = RandomForestClassifier(n_estimators=n_estimators, max_features = max_features, max_depth=max_depth, min_samples_split = min_samples_split, min_samples_leaf = min_samples_leaf, class_weight = "balanced", n_jobs=-1, random_state = seed) # get object
    
    # setup pipeline to take PCA, then fit a clf model
    clf_pipe = Pipeline(
        [('minMaxScaler', MinMaxScaler()),
         ('PCA', PCA),
         ('CLF',rfc_clf)]
    )

    accuracy = cross_val_score(clf_pipe, X, y, cv=cv.split(X, y)) # this also can help with parallelism
    MeanAccuracy =  sum(accuracy)/len(accuracy)
    accuracy = np.append(accuracy, MeanAccuracy)
    endTime = datetime.now()
    TotalTime = endTime - startTime
    accuracy = np.append(accuracy, TotalTime)
    
    #print(TotalTime)
    #print(accuracy)
    
    return accuracy
"""

CPU times: user 3 µs, sys: 1e+03 ns, total: 4 µs
Wall time: 8.11 µs

'\ndef rfc_explor_w_PCA(n_estimators,\n               max_features,\n               max_depth, \n               min_samples_split,\n               min_samples_leaf,\n               PCA,\n               Data        = OPMAnalysisDataNoFam,\n               cv          = cv,\n               seed        = seed):\n    startTime = datetime.now()\n    y = Data["SEP"].values # get the labels we want    \n    \n    X = Data.drop("SEP", axis=1).as_matrix()\n    \n    rfc_clf = RandomForestClassifier(n_estimators=n_estimators, max_features = max_features, max_depth=max_depth, min_samples_split = min_samples_split, min_samples_leaf = min_samples_leaf, class_weight = "balanced", n_jobs=-1, random_state = seed) # get object\n    \n    # setup pipeline to take PCA, then fit a clf model\n    clf_pipe = Pipeline(\n        [(\'minMaxScaler\', MinMaxScaler()),\n         (\'PCA\', PCA),\n         (\'CLF\',rfc_clf)]\n    )\n\n    accuracy = cross_val_score(clf_pipe, X, y, cv=cv.split(X, y)) # this also can help with parallelism\n    MeanAccuracy =  sum(accuracy)/len(accuracy)\n    accuracy = np.append(accuracy, MeanAccuracy)\n    endTime = datetime.now()\n    TotalTime = endTime - startTime\n    accuracy = np.append(accuracy, TotalTime)\n    \n    #print(TotalTime)\n    #print(accuracy)\n    \n    return accuracy\n'

We have created a function to be re-used for our cross-validation Accuracy Scores. Inputs of PCA components, Model CLF object, original sample data, and a CV containing our test/train splits allow us to easily produce an array of Accuracy Scores for the different permutations of models tested. A XXXXXXTBDXXXXX plot is also displayed depicting a view of the misclassification values for each iteration. Finally, a confusion matrix is displayed for the last test/train iteration for further interpretation on results.

%%time
"""
acclist = [] 
fullColumns = list(OPMAnalysisDataNoFam.columns)

for i in fullColumns:
    if i == "SEP": fullColumns.remove(i)

n_estimators       =  [10    , 10     , 10    , 10    , 10    , 10    , 10    , 10    , 10    , 10    , 10  , 5    , 15   ]  
max_features       =  ['auto', 'auto' , 'auto', 'auto', 'auto', 'auto', 'auto', 14    , 14    , 14    , 14  , 14   , 14   ] 
max_depth          =  [None  , None   , None  , None  , None  , None  , None  , None  , 1000  , 500   , 100 , 1000 , 1000 ] 
min_samples_split  =  [2     , 8      , 12    , 16    , 20    , 50    , 80    , 50    , 50    , 50    , 50  , 50   , 50   ] 
min_samples_leaf   =  [1     , 4      , 6     , 8     , 10    , 25    , 40    , 25    , 25    , 25    , 25  , 25   , 25   ]

##Model with all Raw Scaled Features
for i in range(0,len(n_estimators)):
    acclist.append(rfc_explor(n_estimators      = n_estimators[i],
                              max_features      = max_features[i],
                              max_depth         = max_depth[i],
                              min_samples_split = min_samples_split[i],
                              min_samples_leaf  = min_samples_leaf[i],
                              cols              = fullColumns
                             )
                  )

rfcdf = pd.DataFrame(pd.concat([pd.DataFrame({  "ModelVersion": "All Raw Features",
                                                "n_estimators": n_estimators,          
                                                "max_features": max_features,         
                                                "max_depth": max_depth,        
                                                "min_samples_split": min_samples_split,
                                                "min_samples_leaf": min_samples_leaf   
                                              }),
                               pd.DataFrame(acclist)], axis = 1).reindex())
rfcdf.columns = ['ModelVersion', 'max_depth', 'max_features', 'min_samples_leaf','min_samples_split', 'n_estimators', 'Iteration 0', 'Iteration 1', 'Iteration 2', 'Iteration 3', 'Iteration 4', 'MeanAccuracy', 'RunTime']
display(rfcdf)
del rfcdf, acclist

acclist = []

n_estimators       =  [10    , 10     , 10    , 10    , 10    , 10    , 10    , 10    , 10    , 10    , 10  , 5    , 15   ]  
max_features       =  ['auto', 'auto' , 'auto', 'auto', 'auto', 'auto', 'auto', 14    , 14    , 14    , 14  , 14   , 14   ] 
max_depth          =  [None  , None   , None  , None  , None  , None  , None  , None  , 1000  , 500   , 100 , 1000 , 1000 ] 
min_samples_split  =  [2     , 8      , 12    , 16    , 20    , 50    , 80    , 50    , 50    , 50    , 50  , 50   , 50   ] 
min_samples_leaf   =  [1     , 4      , 6     , 8     , 10    , 25    , 40    , 25    , 25    , 25    , 25  , 25   , 25   ]

## Model with only top 15 raw Scaled Principal Features 
for i in range(0,len(n_estimators)):
    acclist.append(rfc_explor(n_estimators      = n_estimators[i],
                              max_features      = max_features[i],
                              max_depth         = max_depth[i],
                              min_samples_split = min_samples_split[i],
                              min_samples_leaf  = min_samples_leaf[i]
                             )
                  )

rfcdf = pd.DataFrame(pd.concat([pd.DataFrame({  "ModelVersion": "Top 15 Raw from PC",
                                                "n_estimators": n_estimators,          
                                                "max_features": max_features,         
                                                "max_depth": max_depth,        
                                                "min_samples_split": min_samples_split,
                                                "min_samples_leaf": min_samples_leaf   
                                              }),
                               pd.DataFrame(acclist)], axis = 1).reindex())
rfcdf.columns = ['ModelVersion', 'max_depth', 'max_features', 'min_samples_leaf','min_samples_split', 'n_estimators', 'Iteration 0', 'Iteration 1', 'Iteration 2', 'Iteration 3', 'Iteration 4', 'MeanAccuracy', 'RunTime']
display(rfcdf)
del rfcdf, acclist

### Model with PCA
acclist = []

n_estimators       =  [10    , 10     , 10    , 10    , 10    , 10    , 10    , 10    , 10    , 10    , 10  , 5    , 15   ]  
max_features       =  ['auto', 'auto' , 'auto', 'auto', 'auto', 'auto', 'auto', 14    , 14    , 14    , 14  , 14   , 14   ] 
max_depth          =  [None  , None   , None  , None  , None  , None  , None  , None  , 1000  , 500   , 100 , 1000 , 1000 ] 
min_samples_split  =  [2     , 8      , 12    , 16    , 20    , 50    , 80    , 50    , 50    , 50    , 50  , 50   , 50   ] 
min_samples_leaf   =  [1     , 4      , 6     , 8     , 10    , 25    , 40    , 25    , 25    , 25    , 25  , 25   , 25   ]

for i in range(0,len(n_estimators)):
    acclist.append(rfc_explor_w_PCA(n_estimators      = n_estimators[i],
                                    max_features      = max_features[i],
                                    max_depth         = max_depth[i],
                                    min_samples_split = min_samples_split[i],
                                    min_samples_leaf  = min_samples_leaf[i],
                                    PCA               = PCA(n_components=22, svd_solver='randomized', random_state = seed)
                                   )
                  )

rfcdf = pd.DataFrame(pd.concat([pd.DataFrame({  "ModelVersion": "With PCA",
                                                "n_estimators": n_estimators,          
                                                "max_features": max_features,         
                                                "max_depth": max_depth,        
                                                "min_samples_split": min_samples_split,
                                                "min_samples_leaf": min_samples_leaf   
                                              }),
                               pd.DataFrame(acclist)], axis = 1).reindex())
rfcdf.columns = ['ModelVersion', 'max_depth', 'max_features', 'min_samples_leaf','min_samples_split', 'n_estimators', 'Iteration 0', 'Iteration 1', 'Iteration 2', 'Iteration 3', 'Iteration 4', 'MeanAccuracy', 'RunTime']
display(rfcdf)

#'Iteration 5', 'Iteration 6', 'Iteration 7', 'Iteration 8', 'Iteration 9', 
"""

CPU times: user 3 µs, sys: 1e+03 ns, total: 4 µs
Wall time: 8.11 µs

'\nacclist = [] \nfullColumns = list(OPMAnalysisDataNoFam.columns)\n\nfor i in fullColumns:\n    if i == "SEP": fullColumns.remove(i)\n\nn_estimators       =  [10    , 10     , 10    , 10    , 10    , 10    , 10    , 10    , 10    , 10    , 10  , 5    , 15   ]  \nmax_features       =  [\'auto\', \'auto\' , \'auto\', \'auto\', \'auto\', \'auto\', \'auto\', 14    , 14    , 14    , 14  , 14   , 14   ] \nmax_depth          =  [None  , None   , None  , None  , None  , None  , None  , None  , 1000  , 500   , 100 , 1000 , 1000 ] \nmin_samples_split  =  [2     , 8      , 12    , 16    , 20    , 50    , 80    , 50    , 50    , 50    , 50  , 50   , 50   ] \nmin_samples_leaf   =  [1     , 4      , 6     , 8     , 10    , 25    , 40    , 25    , 25    , 25    , 25  , 25   , 25   ]\n\n##Model with all Raw Scaled Features\nfor i in range(0,len(n_estimators)):\n    acclist.append(rfc_explor(n_estimators      = n_estimators[i],\n                              max_features      = max_features[i],\n                              max_depth         = max_depth[i],\n                              min_samples_split = min_samples_split[i],\n                              min_samples_leaf  = min_samples_leaf[i],\n                              cols              = fullColumns\n                             )\n                  )\n\nrfcdf = pd.DataFrame(pd.concat([pd.DataFrame({  "ModelVersion": "All Raw Features",\n                                                "n_estimators": n_estimators,          \n                                                "max_features": max_features,         \n                                                "max_depth": max_depth,        \n                                                "min_samples_split": min_samples_split,\n                                                "min_samples_leaf": min_samples_leaf   \n                                              }),\n                               pd.DataFrame(acclist)], axis = 1).reindex())\nrfcdf.columns = [\'ModelVersion\', \'max_depth\', \'max_features\', \'min_samples_leaf\',\'min_samples_split\', \'n_estimators\', \'Iteration 0\', \'Iteration 1\', \'Iteration 2\', \'Iteration 3\', \'Iteration 4\', \'MeanAccuracy\', \'RunTime\']\ndisplay(rfcdf)\ndel rfcdf, acclist\n\nacclist = []\n\nn_estimators       =  [10    , 10     , 10    , 10    , 10    , 10    , 10    , 10    , 10    , 10    , 10  , 5    , 15   ]  \nmax_features       =  [\'auto\', \'auto\' , \'auto\', \'auto\', \'auto\', \'auto\', \'auto\', 14    , 14    , 14    , 14  , 14   , 14   ] \nmax_depth          =  [None  , None   , None  , None  , None  , None  , None  , None  , 1000  , 500   , 100 , 1000 , 1000 ] \nmin_samples_split  =  [2     , 8      , 12    , 16    , 20    , 50    , 80    , 50    , 50    , 50    , 50  , 50   , 50   ] \nmin_samples_leaf   =  [1     , 4      , 6     , 8     , 10    , 25    , 40    , 25    , 25    , 25    , 25  , 25   , 25   ]\n\n## Model with only top 15 raw Scaled Principal Features \nfor i in range(0,len(n_estimators)):\n    acclist.append(rfc_explor(n_estimators      = n_estimators[i],\n                              max_features      = max_features[i],\n                              max_depth         = max_depth[i],\n                              min_samples_split = min_samples_split[i],\n                              min_samples_leaf  = min_samples_leaf[i]\n                             )\n                  )\n\nrfcdf = pd.DataFrame(pd.concat([pd.DataFrame({  "ModelVersion": "Top 15 Raw from PC",\n                                                "n_estimators": n_estimators,          \n                                                "max_features": max_features,         \n                                                "max_depth": max_depth,        \n                                                "min_samples_split": min_samples_split,\n                                                "min_samples_leaf": min_samples_leaf   \n                                              }),\n                               pd.DataFrame(acclist)], axis = 1).reindex())\nrfcdf.columns = [\'ModelVersion\', \'max_depth\', \'max_features\', \'min_samples_leaf\',\'min_samples_split\', \'n_estimators\', \'Iteration 0\', \'Iteration 1\', \'Iteration 2\', \'Iteration 3\', \'Iteration 4\', \'MeanAccuracy\', \'RunTime\']\ndisplay(rfcdf)\ndel rfcdf, acclist\n\n### Model with PCA\nacclist = []\n\nn_estimators       =  [10    , 10     , 10    , 10    , 10    , 10    , 10    , 10    , 10    , 10    , 10  , 5    , 15   ]  \nmax_features       =  [\'auto\', \'auto\' , \'auto\', \'auto\', \'auto\', \'auto\', \'auto\', 14    , 14    , 14    , 14  , 14   , 14   ] \nmax_depth          =  [None  , None   , None  , None  , None  , None  , None  , None  , 1000  , 500   , 100 , 1000 , 1000 ] \nmin_samples_split  =  [2     , 8      , 12    , 16    , 20    , 50    , 80    , 50    , 50    , 50    , 50  , 50   , 50   ] \nmin_samples_leaf   =  [1     , 4      , 6     , 8     , 10    , 25    , 40    , 25    , 25    , 25    , 25  , 25   , 25   ]\n\nfor i in range(0,len(n_estimators)):\n    acclist.append(rfc_explor_w_PCA(n_estimators      = n_estimators[i],\n                                    max_features      = max_features[i],\n                                    max_depth         = max_depth[i],\n                                    min_samples_split = min_samples_split[i],\n                                    min_samples_leaf  = min_samples_leaf[i],\n                                    PCA               = PCA(n_components=22, svd_solver=\'randomized\', random_state = seed)\n                                   )\n                  )\n\nrfcdf = pd.DataFrame(pd.concat([pd.DataFrame({  "ModelVersion": "With PCA",\n                                                "n_estimators": n_estimators,          \n                                                "max_features": max_features,         \n                                                "max_depth": max_depth,        \n                                                "min_samples_split": min_samples_split,\n                                                "min_samples_leaf": min_samples_leaf   \n                                              }),\n                               pd.DataFrame(acclist)], axis = 1).reindex())\nrfcdf.columns = [\'ModelVersion\', \'max_depth\', \'max_features\', \'min_samples_leaf\',\'min_samples_split\', \'n_estimators\', \'Iteration 0\', \'Iteration 1\', \'Iteration 2\', \'Iteration 3\', \'Iteration 4\', \'MeanAccuracy\', \'RunTime\']\ndisplay(rfcdf)\n\n#\'Iteration 5\', \'Iteration 6\', \'Iteration 7\', \'Iteration 8\', \'Iteration 9\', \n'

"""
def plot_confusion_matrix(cm, classes,
                          normalize=False,
                          title='Confusion matrix',
                          cmap=plt.cm.Blues):
    
    #This function prints and plots the confusion matrix.
    #Normalization can be applied by setting `normalize=True`.
    
    plt.rcParams['figure.figsize'] = (18, 6)
    plt.rcParams.update({'font.size': 16})
    plt.rc('xtick', labelsize=18)
    plt.rc('ytick', labelsize=18) 
    plt.imshow(cm, interpolation='nearest', cmap=cmap)
    plt.title(title, fontsize = 18)
    plt.colorbar()
    tick_marks = np.arange(len(classes))
    plt.xticks(tick_marks, classes, rotation=45)
    plt.yticks(tick_marks, classes)

    if normalize:
        cm = cm.astype('float') / cm.sum(axis=1)[:, np.newaxis]
        print("Normalized confusion matrix")
    else:
        print('Confusion matrix, without normalization')

    print(cm)

    thresh = cm.max() / 2.
    for i, j in itertools.product(range(cm.shape[0]), range(cm.shape[1])):
        plt.text(j, i, round(cm[i, j],2),
                 horizontalalignment="center",
                 color="white" if cm[i, j] > thresh else "black")

    plt.tight_layout()
    plt.ylabel('True label', fontsize = 18)
    plt.xlabel('Predicted label', fontsize = 18)

    plt.show()
"""

'\ndef plot_confusion_matrix(cm, classes,\n                          normalize=False,\n                          title=\'Confusion matrix\',\n                          cmap=plt.cm.Blues):\n    \n    #This function prints and plots the confusion matrix.\n    #Normalization can be applied by setting `normalize=True`.\n    \n    plt.rcParams[\'figure.figsize\'] = (18, 6)\n    plt.rcParams.update({\'font.size\': 16})\n    plt.rc(\'xtick\', labelsize=18)\n    plt.rc(\'ytick\', labelsize=18) \n    plt.imshow(cm, interpolation=\'nearest\', cmap=cmap)\n    plt.title(title, fontsize = 18)\n    plt.colorbar()\n    tick_marks = np.arange(len(classes))\n    plt.xticks(tick_marks, classes, rotation=45)\n    plt.yticks(tick_marks, classes)\n\n    if normalize:\n        cm = cm.astype(\'float\') / cm.sum(axis=1)[:, np.newaxis]\n        print("Normalized confusion matrix")\n    else:\n        print(\'Confusion matrix, without normalization\')\n\n    print(cm)\n\n    thresh = cm.max() / 2.\n    for i, j in itertools.product(range(cm.shape[0]), range(cm.shape[1])):\n        plt.text(j, i, round(cm[i, j],2),\n                 horizontalalignment="center",\n                 color="white" if cm[i, j] > thresh else "black")\n\n    plt.tight_layout()\n    plt.ylabel(\'True label\', fontsize = 18)\n    plt.xlabel(\'Predicted label\', fontsize = 18)\n\n    plt.show()\n'

%%time
"""
def compute_kfold_scores_Classification( clf,
                                         Data     = OPMAnalysisDataNoFam,
                                         cols     = PCList,
                                         cv       = cv):

    y = Data["SEP"].values # get the labels we want    
    
    y = np.where(y == 'NS', 0, 
                 np.where(y == 'SA', 1,
                          np.where(y == 'SC', 2,
                                   np.where(y == 'SD', 3,
                                            np.where(y == 'SH', 4,
                                                     5
                                                    )
                                           )
                                  )
                         )
                )
    
    X = Data[cols].as_matrix()


    # Run classifier with cross-validation and plot ROC curves

    # setup pipeline to take PCA, then fit a clf model
    clf_pipe = Pipeline(
        [('minMaxScaler', MinMaxScaler()),
         ('CLF',clf)]
    )
    
    
    accuracy = []
    #logloss = []
    
    for (train, test), color in zip(cv.split(X, y), colors):
        clf_pipe.fit(X[train],y[train])  # train object
        y_hat = clf_pipe.predict(X[test]) # get test set preditions
        
        a = float(mt.accuracy_score(y[test],y_hat))
        #l = float(mt.log_loss(y[test], y_hat))
        
        accuracy.append(round(a,5)) 

        #logloss.append(round(l,5)) 
    
    #print("Accuracy Ratings across all iterations: {0}\n\n\
#Average Accuracy: {1}\n\n\
#Log Loss Values across all iterations: {2}\n\n\
#Average Log Loss: {3}\n".format(accuracy, round(sum(accuracy)/len(accuracy),5), logloss,round(sum(logloss)/len(logloss),5)))

    print("Accuracy Ratings across all iterations: {0}\n\n\
Average Accuracy: {1}\n".format(accuracy, round(sum(accuracy)/len(accuracy),5)))

    
    ytestnames = np.where(y[test] ==  0,'NS', 
                          np.where(y[test] ==  1,'SA',
                                   np.where(y[test] ==  2,'SC',
                                            np.where(y[test] ==  3,'SD',
                                                     np.where(y[test] ==  4,'SH',
                                                              'SI'
                                                             )
                                                    )
                                           )
                                  )
                         )
    
    yhatnames  = np.where(y_hat ==  0,'NS', 
                          np.where(y_hat ==  1,'SA',
                                   np.where(y_hat ==  2,'SC',
                                            np.where(y_hat ==  3,'SD',
                                                     np.where(y_hat ==  4,'SH',
                                                              'SI'
                                                             )
                                                    )
                                           )
                                  )
                         )
    #print(set(list(y_hat)))
    print("confusion matrix\n{0}\n".format(pd.crosstab(ytestnames, yhatnames, rownames = ['True'], colnames = ['Predicted'], margins = True)))
        
        # Plot non-normalized confusion matrix
    plt.figure()
    plot_confusion_matrix(confusion_matrix(y[test], y_hat), 
                          classes   =["NS",  "SA",   "SC", "SD",  "SI"], 
                          normalize =True,
                          title     ='Confusion matrix, with normalization')
    
    return clf_pipe.named_steps['CLF'], accuracy
"""

CPU times: user 3 µs, sys: 1e+03 ns, total: 4 µs
Wall time: 8.34 µs

'\ndef compute_kfold_scores_Classification( clf,\n                                         Data     = OPMAnalysisDataNoFam,\n                                         cols     = PCList,\n                                         cv       = cv):\n\n    y = Data["SEP"].values # get the labels we want    \n    \n    y = np.where(y == \'NS\', 0, \n                 np.where(y == \'SA\', 1,\n                          np.where(y == \'SC\', 2,\n                                   np.where(y == \'SD\', 3,\n                                            np.where(y == \'SH\', 4,\n                                                     5\n                                                    )\n                                           )\n                                  )\n                         )\n                )\n    \n    X = Data[cols].as_matrix()\n\n\n    # Run classifier with cross-validation and plot ROC curves\n\n    # setup pipeline to take PCA, then fit a clf model\n    clf_pipe = Pipeline(\n        [(\'minMaxScaler\', MinMaxScaler()),\n         (\'CLF\',clf)]\n    )\n    \n    \n    accuracy = []\n    #logloss = []\n    \n    for (train, test), color in zip(cv.split(X, y), colors):\n        clf_pipe.fit(X[train],y[train])  # train object\n        y_hat = clf_pipe.predict(X[test]) # get test set preditions\n        \n        a = float(mt.accuracy_score(y[test],y_hat))\n        #l = float(mt.log_loss(y[test], y_hat))\n        \n        accuracy.append(round(a,5)) \n\n        #logloss.append(round(l,5)) \n    \n    #print("Accuracy Ratings across all iterations: {0}\n\n#Average Accuracy: {1}\n\n#Log Loss Values across all iterations: {2}\n\n#Average Log Loss: {3}\n".format(accuracy, round(sum(accuracy)/len(accuracy),5), logloss,round(sum(logloss)/len(logloss),5)))\n\n    print("Accuracy Ratings across all iterations: {0}\n\nAverage Accuracy: {1}\n".format(accuracy, round(sum(accuracy)/len(accuracy),5)))\n\n    \n    ytestnames = np.where(y[test] ==  0,\'NS\', \n                          np.where(y[test] ==  1,\'SA\',\n                                   np.where(y[test] ==  2,\'SC\',\n                                            np.where(y[test] ==  3,\'SD\',\n                                                     np.where(y[test] ==  4,\'SH\',\n                                                              \'SI\'\n                                                             )\n                                                    )\n                                           )\n                                  )\n                         )\n    \n    yhatnames  = np.where(y_hat ==  0,\'NS\', \n                          np.where(y_hat ==  1,\'SA\',\n                                   np.where(y_hat ==  2,\'SC\',\n                                            np.where(y_hat ==  3,\'SD\',\n                                                     np.where(y_hat ==  4,\'SH\',\n                                                              \'SI\'\n                                                             )\n                                                    )\n                                           )\n                                  )\n                         )\n    #print(set(list(y_hat)))\n    print("confusion matrix\n{0}\n".format(pd.crosstab(ytestnames, yhatnames, rownames = [\'True\'], colnames = [\'Predicted\'], margins = True)))\n        \n        # Plot non-normalized confusion matrix\n    plt.figure()\n    plot_confusion_matrix(confusion_matrix(y[test], y_hat), \n                          classes   =["NS",  "SA",   "SC", "SD",  "SI"], \n                          normalize =True,\n                          title     =\'Confusion matrix, with normalization\')\n    \n    return clf_pipe.named_steps[\'CLF\'], accuracy\n'

%%time
"""
rfc_clf = RandomForestClassifier(n_estimators       = 15, 
                                 max_features       = 14, 
                                 max_depth          = 1000.0, 
                                 min_samples_split  = 50, 
                                 min_samples_leaf   = 25, 
                                 class_weight       = "balanced",
                                 n_jobs             = -1, 
                                 random_state       = seed) # get object
    
rfc_clf, rfc_acc = compute_kfold_scores_Classification(rfc_clf, cols = fullColumns)
"""

CPU times: user 3 µs, sys: 1e+03 ns, total: 4 µs
Wall time: 8.11 µs

'\nrfc_clf = RandomForestClassifier(n_estimators       = 15, \n                                 max_features       = 14, \n                                 max_depth          = 1000.0, \n                                 min_samples_split  = 50, \n                                 min_samples_leaf   = 25, \n                                 class_weight       = "balanced",\n                                 n_jobs             = -1, \n                                 random_state       = seed) # get object\n    \nrfc_clf, rfc_acc = compute_kfold_scores_Classification(rfc_clf, cols = fullColumns)\n'

"""
list(OPMAnalysisDataNoFam.SEP.unique())
"""

'\nlist(OPMAnalysisDataNoFam.SEP.unique())\n'

**Experimenting with multiclass ROC curves below.**

'''%%time

def compute_kfold_scores_Classification( clf,
                                         Data     = OPMAnalysisDataNoFam,
                                         cols     = PCList,
                                         cv       = cv):

    y = Data["SEP"].values # get the labels we want    
    
    y = np.where(y == 'NS', 0, 
                 np.where(y == 'SA', 1,
                          np.where(y == 'SC', 2,
                                   np.where(y == 'SD', 3,
                                            np.where(y == 'SH', 4,
                                                     5
                                                    )
                                           )
                                  )
                         )
                )
    
    X = Data[cols].as_matrix()
    
    # Binarize the output
    y_bin = label_binarize(Data["SEP"].values, list(Data.SEP.unique()))
    n_classes = y_bin.shape[1]

    # Run classifier with cross-validation and plot ROC curves

    # setup pipeline to take PCA, then fit a clf model
    clf_pipe = Pipeline(
        [('minMaxScaler', MinMaxScaler()),
         ('CLF',clf)]
    )
    
    colors = cycle(['cyan', 'indigo', 'seagreen', 'yellow', 'blue', 'darkorange', 'pink', 'darkred', 'dimgray', 'maroon', 'coral'])

    accuracy = []
    #logloss = []
    
    for (train, test), color in zip(cv.split(X, y), colors):
        clf_pipe.fit(X[train],y[train])  # train object
        y_hat = clf_pipe.predict(X[test]) # get test set preditions
        
        a = float(mt.accuracy_score(y[test],y_hat))
        #l = float(mt.log_loss(y[test], y_hat))
        
        accuracy.append(round(a,5)) 

        #logloss.append(round(l,5))
        
        # Compute ROC curve and area the curve
        #fpr, tpr, thresholds = roc_curve(y[test], probas_[:, 1])
        #mean_tpr += interp(mean_fpr, fpr, tpr)
        #mean_tpr[0] = 0.0
        #roc_auc = auc(fpr, tpr)
        #
        #plt.rcParams['figure.figsize'] = (12, 6)
        #
        #plt.plot(fpr, tpr, lw=lw, color=color,
        #         label='ROC fold %d (area = %0.2f)' % (i, roc_auc))
#
        #i += 1
        # Compute ROC curve and ROC area for each class
        fpr = dict()
        tpr = dict()
        roc_auc = dict()
        for i in range(n_classes):
            fpr[i], tpr[i], _ = roc_curve(y[test][:, i], y_hat[:, i])
            roc_auc[i] = auc(fpr[i], tpr[i])

        # Plot of a ROC curve for a specific class
        plt.figure()
        lw = 2
        plt.plot(fpr[2], tpr[2], color='darkorange',
                 lw=lw, label='ROC curve (area = %0.2f)' % roc_auc[2])
        plt.plot([0, 1], [0, 1], color='navy', lw=lw, linestyle='--')
        plt.xlim([0.0, 1.0])
        plt.ylim([0.0, 1.05])
        plt.xlabel('False Positive Rate')
        plt.ylabel('True Positive Rate')
        plt.title('Receiver operating characteristic example')
        plt.legend(loc="lower right")
        plt.show()
    #print("Accuracy Ratings across all iterations: {0}\n\n\
#Average Accuracy: {1}\n\n\
#Log Loss Values across all iterations: {2}\n\n\
#Average Log Loss: {3}\n".format(accuracy, round(sum(accuracy)/len(accuracy),5), logloss,round(sum(logloss)/len(logloss),5)))

    print("Accuracy Ratings across all iterations: {0}\n\n\
Average Accuracy: {1}\n".format(accuracy, round(sum(accuracy)/len(accuracy),5)))

    
    ytestnames = np.where(y[test] ==  0,'NS', 
                          np.where(y[test] ==  1,'SA',
                                   np.where(y[test] ==  2,'SC',
                                            np.where(y[test] ==  3,'SD',
                                                     np.where(y[test] ==  4,'SH',
                                                              'SI'
                                                             )
                                                    )
                                           )
                                  )
                         )
    
    yhatnames  = np.where(y_hat ==  0,'NS', 
                          np.where(y_hat ==  1,'SA',
                                   np.where(y_hat ==  2,'SC',
                                            np.where(y_hat ==  3,'SD',
                                                     np.where(y_hat ==  4,'SH',
                                                              'SI'
                                                             )
                                                    )
                                           )
                                  )
                         )
    #print(set(list(y_hat)))
    print("confusion matrix\n{0}\n".format(pd.crosstab(ytestnames, yhatnames, rownames = ['True'], colnames = ['Predicted'], margins = True)))
        
        # Plot non-normalized confusion matrix
    plt.figure()
    plot_confusion_matrix(confusion_matrix(y[test], y_hat), 
                          classes   =["NS",  "SA",   "SC", "SD",  "SH",  "SI"], 
                          normalize =True,
                          title     ='Confusion matrix, with normalization')
    
    return clf_pipe.named_steps['CLF'], accuracy'''

'%%time\n\ndef compute_kfold_scores_Classification( clf,\n                                         Data     = OPMAnalysisDataNoFam,\n                                         cols     = PCList,\n                                         cv       = cv):\n\n    y = Data["SEP"].values # get the labels we want    \n    \n    y = np.where(y == \'NS\', 0, \n                 np.where(y == \'SA\', 1,\n                          np.where(y == \'SC\', 2,\n                                   np.where(y == \'SD\', 3,\n                                            np.where(y == \'SH\', 4,\n                                                     5\n                                                    )\n                                           )\n                                  )\n                         )\n                )\n    \n    X = Data[cols].as_matrix()\n    \n    # Binarize the output\n    y_bin = label_binarize(Data["SEP"].values, list(Data.SEP.unique()))\n    n_classes = y_bin.shape[1]\n\n    # Run classifier with cross-validation and plot ROC curves\n\n    # setup pipeline to take PCA, then fit a clf model\n    clf_pipe = Pipeline(\n        [(\'minMaxScaler\', MinMaxScaler()),\n         (\'CLF\',clf)]\n    )\n    \n    colors = cycle([\'cyan\', \'indigo\', \'seagreen\', \'yellow\', \'blue\', \'darkorange\', \'pink\', \'darkred\', \'dimgray\', \'maroon\', \'coral\'])\n\n    accuracy = []\n    #logloss = []\n    \n    for (train, test), color in zip(cv.split(X, y), colors):\n        clf_pipe.fit(X[train],y[train])  # train object\n        y_hat = clf_pipe.predict(X[test]) # get test set preditions\n        \n        a = float(mt.accuracy_score(y[test],y_hat))\n        #l = float(mt.log_loss(y[test], y_hat))\n        \n        accuracy.append(round(a,5)) \n\n        #logloss.append(round(l,5))\n        \n        # Compute ROC curve and area the curve\n        #fpr, tpr, thresholds = roc_curve(y[test], probas_[:, 1])\n        #mean_tpr += interp(mean_fpr, fpr, tpr)\n        #mean_tpr[0] = 0.0\n        #roc_auc = auc(fpr, tpr)\n        #\n        #plt.rcParams[\'figure.figsize\'] = (12, 6)\n        #\n        #plt.plot(fpr, tpr, lw=lw, color=color,\n        #         label=\'ROC fold %d (area = %0.2f)\' % (i, roc_auc))\n#\n        #i += 1\n        # Compute ROC curve and ROC area for each class\n        fpr = dict()\n        tpr = dict()\n        roc_auc = dict()\n        for i in range(n_classes):\n            fpr[i], tpr[i], _ = roc_curve(y[test][:, i], y_hat[:, i])\n            roc_auc[i] = auc(fpr[i], tpr[i])\n\n        # Plot of a ROC curve for a specific class\n        plt.figure()\n        lw = 2\n        plt.plot(fpr[2], tpr[2], color=\'darkorange\',\n                 lw=lw, label=\'ROC curve (area = %0.2f)\' % roc_auc[2])\n        plt.plot([0, 1], [0, 1], color=\'navy\', lw=lw, linestyle=\'--\')\n        plt.xlim([0.0, 1.0])\n        plt.ylim([0.0, 1.05])\n        plt.xlabel(\'False Positive Rate\')\n        plt.ylabel(\'True Positive Rate\')\n        plt.title(\'Receiver operating characteristic example\')\n        plt.legend(loc="lower right")\n        plt.show()\n    #print("Accuracy Ratings across all iterations: {0}\n\n#Average Accuracy: {1}\n\n#Log Loss Values across all iterations: {2}\n\n#Average Log Loss: {3}\n".format(accuracy, round(sum(accuracy)/len(accuracy),5), logloss,round(sum(logloss)/len(logloss),5)))\n\n    print("Accuracy Ratings across all iterations: {0}\n\nAverage Accuracy: {1}\n".format(accuracy, round(sum(accuracy)/len(accuracy),5)))\n\n    \n    ytestnames = np.where(y[test] ==  0,\'NS\', \n                          np.where(y[test] ==  1,\'SA\',\n                                   np.where(y[test] ==  2,\'SC\',\n                                            np.where(y[test] ==  3,\'SD\',\n                                                     np.where(y[test] ==  4,\'SH\',\n                                                              \'SI\'\n                                                             )\n                                                    )\n                                           )\n                                  )\n                         )\n    \n    yhatnames  = np.where(y_hat ==  0,\'NS\', \n                          np.where(y_hat ==  1,\'SA\',\n                                   np.where(y_hat ==  2,\'SC\',\n                                            np.where(y_hat ==  3,\'SD\',\n                                                     np.where(y_hat ==  4,\'SH\',\n                                                              \'SI\'\n                                                             )\n                                                    )\n                                           )\n                                  )\n                         )\n    #print(set(list(y_hat)))\n    print("confusion matrix\n{0}\n".format(pd.crosstab(ytestnames, yhatnames, rownames = [\'True\'], colnames = [\'Predicted\'], margins = True)))\n        \n        # Plot non-normalized confusion matrix\n    plt.figure()\n    plot_confusion_matrix(confusion_matrix(y[test], y_hat), \n                          classes   =["NS",  "SA",   "SC", "SD",  "SH",  "SI"], \n                          normalize =True,\n                          title     =\'Confusion matrix, with normalization\')\n    \n    return clf_pipe.named_steps[\'CLF\'], accuracy'

'''%%time

rfc_clf = OneVsRestClassifier(RandomForestClassifier(n_estimators       = 15, 
                                 max_features       = 14, 
                                 max_depth          = 1000.0, 
                                 min_samples_split  = 50, 
                                 min_samples_leaf   = 25, 
                                 class_weight       = "balanced",
                                 n_jobs             = -1, 
                                 random_state       = seed)) # get object
    
rfc_clf, rfc_acc = compute_kfold_scores_Classification(rfc_clf, cols = fullColumns)'''

'%%time\n\nrfc_clf = OneVsRestClassifier(RandomForestClassifier(n_estimators       = 15, \n                                 max_features       = 14, \n                                 max_depth          = 1000.0, \n                                 min_samples_split  = 50, \n                                 min_samples_leaf   = 25, \n                                 class_weight       = "balanced",\n                                 n_jobs             = -1, \n                                 random_state       = seed)) # get object\n    \nrfc_clf, rfc_acc = compute_kfold_scores_Classification(rfc_clf, cols = fullColumns)'

New attempt:¶

'''%%time

rfc_clf = OneVsRestClassifier(RandomForestClassifier(n_estimators       = 15, 
                                 max_features       = 14, 
                                 max_depth          = 1000.0, 
                                 min_samples_split  = 50, 
                                 min_samples_leaf   = 25, 
                                 class_weight       = "balanced",
                                 n_jobs             = -1, 
                                 random_state       = seed)) # get object

y = OPMAnalysisDataNoFam["SEP"].values # get the labels we want
    
y = np.where(y == 'NS', 0, 
             np.where(y == 'SA', 1,
                      np.where(y == 'SC', 2,
                               np.where(y == 'SD', 3,
                                        np.where(y == 'SH', 4,
                                                 5
                                                )
                                       )
                              )
                     )
            )

X = OPMAnalysisDataNoFam[fullColumns].as_matrix()

# Binarize the output
#y_bin = label_binarize(OPMAnalysisDataNoFam["SEP"].values, list(OPMAnalysisDataNoFam.SEP.unique()))
n_classes = len(set(y))

#classifier = OneVsRestClassifier(svm.SVC(kernel='linear', probability=True,
#                                 random_state=random_state))
#y_score = rfc_clf.fit(X[train], y[train]).decision_function(X_test)
#
colors = cycle(['cyan', 'indigo', 'seagreen', 'yellow', 'blue', 'darkorange', 'pink', 'darkred', 'dimgray', 'maroon', 'coral'])
#
#accuracy = []
#

#for (train, test), color in zip(cv.split(X, y), colors):
#    probas_ = rfc_clf.fit(X[train], y[train]).predict_proba(X[test])    
#    
#    #rfc_clf.fit(X[train],y[train])  # train object
#    y_hat = rfc_clf.predict(X[test]) # get test set preditions
#    #y_hat = rfc_clf.fit(X[train],y[train]).decision_function(X[test])
#    
#    fpr = dict()
#    tpr = dict()
#    roc_auc = dict()
#    for i in range(n_classes):
#        #fpr[i], tpr[i], _ = roc_curve(y[test][:, i], y_hat[:, i])
#        print(len(probas_[:, i]))
#        print(y[test])
#        #fpr[i], tpr[i], thresholds = roc_curve(y[test][:, i], probas_[:, i])
#        #roc_auc[i] = auc(fpr[i], tpr[i])

X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=.5, random_state=seed)

probas_ = rfc_clf.fit(X_train, y_train).predict_proba(X_test)    
    
#rfc_clf.fit(X[train],y[train])  # train object
#y_hat = rfc_clf.predict(X_test) # get test set preditions
#y_hat = rfc_clf.fit(X_train,y_train).decision_function(X_test)

fpr = dict()
tpr = dict()
roc_auc = dict()
for i in range(n_classes):
    #fpr[i], tpr[i], _ = roc_curve(y[test][:, i], y_hat[:, i])
    print(probas_[:, i])
    print(y_test)
    #fpr[i], tpr[i], thresholds = roc_curve(y[test][i], probas_[:, i])
    #roc_auc[i] = auc(fpr[i], tpr[i])'''

'%%time\n\nrfc_clf = OneVsRestClassifier(RandomForestClassifier(n_estimators       = 15, \n                                 max_features       = 14, \n                                 max_depth          = 1000.0, \n                                 min_samples_split  = 50, \n                                 min_samples_leaf   = 25, \n                                 class_weight       = "balanced",\n                                 n_jobs             = -1, \n                                 random_state       = seed)) # get object\n\ny = OPMAnalysisDataNoFam["SEP"].values # get the labels we want\n    \ny = np.where(y == \'NS\', 0, \n             np.where(y == \'SA\', 1,\n                      np.where(y == \'SC\', 2,\n                               np.where(y == \'SD\', 3,\n                                        np.where(y == \'SH\', 4,\n                                                 5\n                                                )\n                                       )\n                              )\n                     )\n            )\n\nX = OPMAnalysisDataNoFam[fullColumns].as_matrix()\n\n# Binarize the output\n#y_bin = label_binarize(OPMAnalysisDataNoFam["SEP"].values, list(OPMAnalysisDataNoFam.SEP.unique()))\nn_classes = len(set(y))\n\n#classifier = OneVsRestClassifier(svm.SVC(kernel=\'linear\', probability=True,\n#                                 random_state=random_state))\n#y_score = rfc_clf.fit(X[train], y[train]).decision_function(X_test)\n#\ncolors = cycle([\'cyan\', \'indigo\', \'seagreen\', \'yellow\', \'blue\', \'darkorange\', \'pink\', \'darkred\', \'dimgray\', \'maroon\', \'coral\'])\n#\n#accuracy = []\n#\n\n#for (train, test), color in zip(cv.split(X, y), colors):\n#    probas_ = rfc_clf.fit(X[train], y[train]).predict_proba(X[test])    \n#    \n#    #rfc_clf.fit(X[train],y[train])  # train object\n#    y_hat = rfc_clf.predict(X[test]) # get test set preditions\n#    #y_hat = rfc_clf.fit(X[train],y[train]).decision_function(X[test])\n#    \n#    fpr = dict()\n#    tpr = dict()\n#    roc_auc = dict()\n#    for i in range(n_classes):\n#        #fpr[i], tpr[i], _ = roc_curve(y[test][:, i], y_hat[:, i])\n#        print(len(probas_[:, i]))\n#        print(y[test])\n#        #fpr[i], tpr[i], thresholds = roc_curve(y[test][:, i], probas_[:, i])\n#        #roc_auc[i] = auc(fpr[i], tpr[i])\n\nX_train, X_test, y_train, y_test = train_test_split(X, y, test_size=.5, random_state=seed)\n\nprobas_ = rfc_clf.fit(X_train, y_train).predict_proba(X_test)    \n    \n#rfc_clf.fit(X[train],y[train])  # train object\n#y_hat = rfc_clf.predict(X_test) # get test set preditions\n#y_hat = rfc_clf.fit(X_train,y_train).decision_function(X_test)\n\nfpr = dict()\ntpr = dict()\nroc_auc = dict()\nfor i in range(n_classes):\n    #fpr[i], tpr[i], _ = roc_curve(y[test][:, i], y_hat[:, i])\n    print(probas_[:, i])\n    print(y_test)\n    #fpr[i], tpr[i], thresholds = roc_curve(y[test][i], probas_[:, i])\n    #roc_auc[i] = auc(fpr[i], tpr[i])'

Building a Binary Classification model¶

Reducing data to Non-Separation and Quit Separation Types¶

Separation Response Weights¶

As was done before, we assess weights across classes. Since stratification was performed previously, we have equal weights. Thus, we can ignore weighting in our binary classifications.

OPMClassWeights = class_weight.compute_class_weight("balanced", OPMAnalysisDataNoFamBinary["SEP"].drop_duplicates(), OPMAnalysisDataNoFamBinary["SEP"])

display(stratumProf.merge(pd.DataFrame({"Weight": OPMClassWeights, "SEP": OPMAnalysisDataNoFamBinary["SEP"].drop_duplicates()}),on="SEP", how="inner"))

Predicting Separation¶

We have chosen to utilize Stratified KFold Cross Validation for our classification analysis, with 5 folds. This means, that from our original sample size of 8002, each "fold" will save off approximately 20% as test observations utilizing the rest as training observations all while keeping the ratio of classes equal amongst customers and subscribers. This process will occur through 5 iterations, or folds, to allow us to cross validate our results amongst different test/train combinations. We have utilized a random_state seed equal to the length of the original sampled dataset to ensure reproducible results.

seed = len(OPMAnalysisDataNoFamBinary)

cv = StratifiedKFold(n_splits = 5, random_state = seed)
print(OPMAnalysisDataNoFamBinary.shape)
print(cv)

(14920, 100)
StratifiedKFold(n_splits=5, random_state=14920, shuffle=False)

Random Forest Classification¶

Max Depth The maximum depth (levels) in the tree. When a value is set, the tree may not split further once this level has been met regardless of how many nodes are in the leaf.

Max Features Number of features to consider when looking for a split.

Minimum Samples in Leaf Minimum number of samples required to be in a leaf node. Splits may not occur which cause the number of samples in a leaf to be less than this value. Too low a value here leads to overfitting the tree to train data.

Minimum Samples to Split Minimum number fo samples required to split a node. Care was taken during parameter tests to keep the ratio between Min Samples in Leaf and Min Samples to Split equal to that of the default values (1:2). This was done to allow an even 50/50 split on nodes which match the lowest granularity split criteria. similar to the min samples in leaf, too low a value here leads to overfitting the tree to train data.

n_estimators Number of Trees generated in the forest. Increasing the number of trees, in our models increased accuracy while decreasing performance. We tuned to provide output that completed all 10 iterations in under 10 minutes.

Not Complete#### After 13 iterations of modifying the above parameters, we land on a final winner based on the highest average Accuracy value across all iterations. Average Accuracy values in our 10 test/train iterations ranged from 70.2668 % from default inputs of the random forest classification model to a value of 72.5192 % in the best tuned model fit. Although the run-time of this model parameter choice is the largest performed, we decided to remain with these inputs due to the amount increase in accuracy. As mentioned previously, we tuned the n_estimators parameter to ensure we stayed under 10 minutes execution. Parameter inputs for the final Random Forest Classification model with the KD Tree Algorithm are as follows: ###Not Complete¶

max_depth	max_features	min_samples_leaf	min_samples_split	n_estimators
TBD	TBD	TBD	TBD	TBD

%%time

def rfc_explorBinary(n_estimators,
               max_features,
               max_depth, 
               min_samples_split,
               min_samples_leaf,
               Data        = OPMAnalysisDataNoFamBinary,
               cols        = PCList,
               cv          = cv,
               seed        = seed):
    startTime = datetime.now()
    y = Data["SEP"].values # get the labels we want    
    
    X = Data[cols].as_matrix()
    
    rfc_clf = RandomForestClassifier(n_estimators=n_estimators, max_features = max_features, max_depth=max_depth, min_samples_split = min_samples_split, min_samples_leaf = min_samples_leaf, n_jobs=-1, random_state = seed) # get object
    
    # setup pipeline to take PCA, then fit a clf model
    clf_pipe = Pipeline(
        [('minMaxScaler', MinMaxScaler()),
         ('CLF',rfc_clf)]
    )

    accuracy = cross_val_score(clf_pipe, X, y, cv=cv.split(X, y)) # this also can help with parallelism
    MeanAccuracy =  sum(accuracy)/len(accuracy)
    accuracy = np.append(accuracy, MeanAccuracy)
    endTime = datetime.now()
    TotalTime = endTime - startTime
    accuracy = np.append(accuracy, TotalTime)
    
    #print(TotalTime)
    #print(accuracy)
    
    return accuracy

CPU times: user 5 µs, sys: 0 ns, total: 5 µs
Wall time: 9.54 µs

%%time

def rfc_explorBinary_w_PCA(n_estimators,
               max_features,
               max_depth, 
               min_samples_split,
               min_samples_leaf,
               PCA,
               Data        = OPMAnalysisDataNoFamBinary,
               cv          = cv,
               seed        = seed):
    startTime = datetime.now()
    y = Data["SEP"].values # get the labels we want    
    
    X = Data.drop("SEP", axis=1).as_matrix()
    
    rfc_clf = RandomForestClassifier(n_estimators=n_estimators, max_features = max_features, max_depth=max_depth, min_samples_split = min_samples_split, min_samples_leaf = min_samples_leaf, n_jobs=-1, random_state = seed) # get object
    
    # setup pipeline to take PCA, then fit a clf model
    clf_pipe = Pipeline(
        [('minMaxScaler', MinMaxScaler()),
         ('PCA', PCA),
         ('CLF',rfc_clf)]
    )

    accuracy = cross_val_score(clf_pipe, X, y, cv=cv.split(X, y)) # this also can help with parallelism
    MeanAccuracy =  sum(accuracy)/len(accuracy)
    accuracy = np.append(accuracy, MeanAccuracy)
    endTime = datetime.now()
    TotalTime = endTime - startTime
    accuracy = np.append(accuracy, TotalTime)
    
    #print(TotalTime)
    #print(accuracy)
    
    return accuracy

CPU times: user 4 µs, sys: 2 µs, total: 6 µs
Wall time: 9.3 µs

%%time
FinalResultsDF = pd.DataFrame(columns= ['ModelVersion', 'Iteration 0', 'Iteration 1', 'Iteration 2', 'Iteration 3', 'Iteration 4', 'MeanAccuracy'])

TopResultsDF = pd.DataFrame(columns= ['ModelVersion', 'Iteration 0', 'Iteration 1', 'Iteration 2', 'Iteration 3', 'Iteration 4', 'MeanAccuracy'])

acclist = [] 
fullColumns = list(OPMAnalysisDataNoFamBinary.columns)

for i in fullColumns:
    if i == "SEP": fullColumns.remove(i)

n_estimators       =  [10    , 10     , 10    , 10    , 10    , 10    , 10    , 10    , 10    , 10    , 10   , 10   , 10    , 10    , 10    , 10    , 10    , 10    , 10    , 10    , 10    , 10    , 10    , 10    , 15    , 20    , 30    , 50    ]  
max_features       =  ['auto', 'auto' , 'auto', 'auto', 'auto', 'auto', 'auto', 'auto', 5     , 10    , 15   , 20   , None  , 15    , 15    , 15    , 15    , 15    , 15    , 15    , 15    , 15    , 15    , 15    , 15    , 15    , 15    , 15    ] 
max_depth          =  [None  , None   , None  , None  , None  , None  , None  , None  , None  , None  , None , None , None  , 10    , 15    , 20    , 25    , 30    , 17    , 18    , 19    , 21    , 22    , 23    , 15    , 15    , 15    , 15    ] 
min_samples_split  =  [2     , 8      , 12    , 18    , 20    , 24    , 36    , 48    , 36    , 36    , 36    , 36  , 36    , 36    , 36    , 36    , 36    , 36    , 36    , 36    , 36    , 36    , 36    , 36    , 36    , 36    , 36    , 36    ] 
min_samples_leaf   =  [1     , 4      , 6     , 9     , 10    , 12    , 18    , 24    , 18    , 18    , 18    , 18  , 18    , 18    , 18    , 18    , 18    , 18    , 18    , 18    , 18    , 18    , 18    , 18    , 18    , 18    , 18    , 18    ]

##Model with all Raw Scaled Features
for i in range(0,len(n_estimators)):
    acclist.append(rfc_explorBinary(n_estimators      = n_estimators[i],
                              max_features      = max_features[i],
                              max_depth         = max_depth[i],
                              min_samples_split = min_samples_split[i],
                              min_samples_leaf  = min_samples_leaf[i],
                              cols              = fullColumns
                             )
                  )

rfcdf = pd.DataFrame(pd.concat([pd.DataFrame({  "ModelVersion": "Random Forest: All Raw Features",
                                                "n_estimators": n_estimators,          
                                                "max_features": max_features,         
                                                "max_depth": max_depth,        
                                                "min_samples_split": min_samples_split,
                                                "min_samples_leaf": min_samples_leaf   
                                              }),
                               pd.DataFrame(acclist)], axis = 1).reindex())
rfcdf.columns = ['ModelVersion', 'max_depth', 'max_features', 'min_samples_leaf','min_samples_split', 'n_estimators', 'Iteration 0', 'Iteration 1', 'Iteration 2', 'Iteration 3', 'Iteration 4', 'MeanAccuracy', 'RunTime']
display(rfcdf)
TopResultsDF = pd.concat([TopResultsDF, rfcdf.sort_values(['MeanAccuracy'], ascending=False)[TopResultsDF.columns].head(1)]).sort_values(['MeanAccuracy'], ascending=False).reset_index(drop=True)
del rfcdf, acclist


acclist = []

n_estimators       =  [10    , 10     , 10    , 10    , 10    , 10    , 10    , 10    , 10    , 10    , 10   , 10   , 10    , 10    , 10    , 10    , 10    , 10    , 10    , 10    , 10    , 10    , 10    , 10    , 15    , 20    , 30    , 50    ]  
max_features       =  ['auto', 'auto' , 'auto', 'auto', 'auto', 'auto', 'auto', 'auto', 5     , 10    , 15   , 20   , None  , 'auto', 'auto', 'auto', 'auto', 'auto', 'auto', 'auto', 'auto', 'auto', 'auto', 'auto', 'auto', 'auto', 'auto', 'auto'] 
max_depth          =  [None  , None   , None  , None  , None  , None  , None  , None  , None  , None  , None , None , None  , 10    , 15    , 20    , 25    , 30    , 17    , 18    , 19    , 21    , 22    , 23    , 17    , 17    , 17    , 17    ] 
min_samples_split  =  [2     , 8      , 12    , 18    , 20    , 24    , 36    , 48    , 48    , 48    , 48   , 48   , 48    , 48    , 48    , 48    , 48    , 48    , 48    , 48    , 48    , 48    , 48    , 48    , 48    , 48    , 48    , 48    ]  
min_samples_leaf   =  [1     , 4      , 6     , 9     , 10    , 12    , 18    , 24    , 24    , 24    , 24   , 24   , 24    , 24    , 24    , 24    , 24    , 24    , 24    , 24    , 24    , 24    , 24    , 24    , 24    , 24    , 24    , 24    ]


## Model with only top 15 raw Scaled Principal Features 
for i in range(0,len(n_estimators)):
    acclist.append(rfc_explorBinary(n_estimators      = n_estimators[i],
                              max_features      = max_features[i],
                              max_depth         = max_depth[i],
                              min_samples_split = min_samples_split[i],
                              min_samples_leaf  = min_samples_leaf[i]
                             )
                  )

rfcdf = pd.DataFrame(pd.concat([pd.DataFrame({  "ModelVersion": "Random Forest: Top 15 Raw from PC",
                                                "n_estimators": n_estimators,          
                                                "max_features": max_features,         
                                                "max_depth": max_depth,        
                                                "min_samples_split": min_samples_split,
                                                "min_samples_leaf": min_samples_leaf   
                                              }),
                               pd.DataFrame(acclist)], axis = 1).reindex())
rfcdf.columns = ['ModelVersion', 'max_depth', 'max_features', 'min_samples_leaf','min_samples_split', 'n_estimators', 'Iteration 0', 'Iteration 1', 'Iteration 2', 'Iteration 3', 'Iteration 4', 'MeanAccuracy', 'RunTime']
display(rfcdf)
TopResultsDF = pd.concat([TopResultsDF, rfcdf.sort_values(['MeanAccuracy'], ascending=False)[TopResultsDF.columns].head(1)]).sort_values(['MeanAccuracy'], ascending=False).reset_index(drop=True)
del rfcdf, acclist


### Model with PCA
acclist = []

n_estimators       =  [10    , 10     , 10    , 10    , 10    , 10    , 10    , 10    , 10    , 10   , 10   , 10    , 10    , 10    , 10    , 10    , 10    , 10    , 10    , 10    , 10    , 10    , 10    , 15    , 20    , 30    , 50    ]  
max_features       =  ['auto', 'auto' , 'auto', 'auto', 'auto', 'auto', 'auto', 5     , 10    , 15   , 20   , None  , 'auto', 'auto', 'auto', 'auto', 'auto', 'auto', 'auto', 'auto', 'auto', 'auto', 'auto', 'auto', 'auto', 'auto', 'auto'] 
max_depth          =  [None  , None   , None  , None  , None  , None  , None  , None  , None  , None , None , None  , 10    , 15    , 20    , 25    , 30    , 17    , 18    , 19    , 21    , 22    , 23    , 15    , 15    , 15    , 15    ] 
min_samples_split  =  [2     , 8      , 12    , 18    , 20    , 24    , 36    ,  24   , 24    , 24    , 24    , 24  , 24    , 24    , 24    , 24    , 24    , 24    , 24    , 24    , 24    , 24    , 24    , 24    , 24    , 24    , 24    ]   
min_samples_leaf   =  [1     , 4      , 6     , 9     , 10    , 12    , 18    ,  12   , 12    , 12    , 12    , 12  , 12    , 12    , 12    , 12    , 12    , 12    , 12    , 12    , 12    , 12    , 12    , 12    , 12    , 12    , 12    ]       


for i in range(0,len(n_estimators)):
    acclist.append(rfc_explorBinary_w_PCA(n_estimators      = n_estimators[i],
                                    max_features      = max_features[i],
                                    max_depth         = max_depth[i],
                                    min_samples_split = min_samples_split[i],
                                    min_samples_leaf  = min_samples_leaf[i],
                                    PCA               = PCA(n_components=22, svd_solver='randomized', random_state = seed)
                                   )
                  )

rfcdf = pd.DataFrame(pd.concat([pd.DataFrame({  "ModelVersion": "Random Forest: With PCA",
                                                "n_estimators": n_estimators,          
                                                "max_features": max_features,         
                                                "max_depth": max_depth,        
                                                "min_samples_split": min_samples_split,
                                                "min_samples_leaf": min_samples_leaf   
                                              }),
                               pd.DataFrame(acclist)], axis = 1).reindex())
rfcdf.columns = ['ModelVersion', 'max_depth', 'max_features', 'min_samples_leaf','min_samples_split', 'n_estimators', 'Iteration 0', 'Iteration 1', 'Iteration 2', 'Iteration 3', 'Iteration 4', 'MeanAccuracy', 'RunTime']

display(rfcdf)
TopResultsDF = pd.concat([TopResultsDF, rfcdf.sort_values(['MeanAccuracy'], ascending=False)[TopResultsDF.columns].head(1)]).sort_values(['MeanAccuracy'], ascending=False).reset_index(drop=True)
del rfcdf, acclist

CPU times: user 5min 59s, sys: 10min 3s, total: 16min 3s
Wall time: 2min 18s

display(TopResultsDF)

plot = TopResultsDF[["Iteration 0","Iteration 1","Iteration 2","Iteration 3","Iteration 4"]].transpose().plot.line(title = "Top Results Among Varying Model Feature Inputs",rot=45)
plot.set_xlabel("Iterations")
plot.set_ylabel("Accuracies")
plot.legend(loc='center left', bbox_to_anchor=(1.01, .5))

FinalResultsDF = pd.concat([FinalResultsDF, TopResultsDF.sort_values(['MeanAccuracy'], ascending=False)[TopResultsDF.columns].head(1)]).sort_values(['MeanAccuracy'], ascending=False).reset_index(drop=True)
TopResultsDF = pd.DataFrame(columns= ['ModelVersion', 'Iteration 0', 'Iteration 1', 'Iteration 2', 'Iteration 3', 'Iteration 4', 'MeanAccuracy'])

/usr/local/es7/lib/python3.5/site-packages/matplotlib/font_manager.py:1297: UserWarning: findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans
  (prop.get_family(), self.defaultFamily[fontext]))

We have created a function to be re-used for our cross-validation Accuracy Scores. Inputs of PCA components, Model CLF object, original sample data, and a CV containing our test/train splits allow us to easily produce an array of Accuracy Scores for the different permutations of models tested. A XXXXXXTBDXXXXX plot is also displayed depicting a view of the misclassification values for each iteration. Finally, a confusion matrix is displayed for the last test/train iteration for further interpretation on results.

def plot_confusion_matrixBinary(cm, classes,
                          normalize=False,
                          title='Confusion matrix',
                          cmap=plt.cm.Blues):
    """
    This function prints and plots the confusion matrix.
    Normalization can be applied by setting `normalize=True`.
    """
    plt.rcParams['figure.figsize'] = (18, 6)
    plt.rcParams.update({'font.size': 16})
    plt.rc('xtick', labelsize=18)
    plt.rc('ytick', labelsize=18) 
    plt.imshow(cm, interpolation='nearest', cmap=cmap)
    plt.title(title, fontsize = 18)
    plt.colorbar()
    tick_marks = np.arange(len(classes))
    plt.xticks(tick_marks, classes, rotation=45)
    plt.yticks(tick_marks, classes)

    if normalize:
        cm = cm.astype('float') / cm.sum(axis=1)[:, np.newaxis]
        print("Normalized confusion matrix")
    else:
        print('Confusion matrix, without normalization')

    print(cm)

    thresh = cm.max() / 2.
    for i, j in itertools.product(range(cm.shape[0]), range(cm.shape[1])):
        plt.text(j, i, round(cm[i, j],2),
                 horizontalalignment="center",
                 color="white" if cm[i, j] > thresh else "black")

    plt.tight_layout()
    plt.ylabel('True label', fontsize = 18)
    plt.xlabel('Predicted label', fontsize = 18)

    plt.show()

def plot_ROC_curve(X, y, mean_tpr, mean_fpr, cv = cv, ):
    
    plt.rcParams['figure.figsize'] = (12, 6)

    lw = 2

    plt.plot([0, 1], [0, 1], linestyle='--', lw=lw, color='k',
             label='Luck')

    mean_tpr /= cv.get_n_splits(X, y)
    mean_tpr[-1] = 1.0
    mean_auc = auc(mean_fpr, mean_tpr)
    plt.plot(mean_fpr, mean_tpr, color='g', linestyle='--',
             label='Mean ROC (area = %0.2f)' % mean_auc, lw=lw)

    plt.xlim([-0.05, 1.05])
    plt.ylim([-0.05, 1.05])
    plt.xlabel('False Positive Rate')
    plt.ylabel('True Positive Rate')
    plt.title('Receiver operating characteristic (ROC) Curve')
    plt.legend(loc="lower right")
    plt.show()

%%time

def compute_kfold_scores_ClassificationBinary( clf,
                                         PCA = "No",
                                         Data     = OPMAnalysisDataNoFamBinary,
                                         cols     = PCList,
                                         cv       = cv):

    y = Data["SEP"].values # get the labels we want    
    
    y = np.where(y == 'NS', 0, 1) # NS = 0; SC = 1
    
    X = Data[cols].as_matrix()


    # Run classifier with cross-validation and plot ROC curves

    # setup pipeline to take PCA, then fit a clf model
    if(PCA == "No"):
        clf_pipe = Pipeline(
            [('minMaxScaler', MinMaxScaler()),
             ('CLF',clf)]
        )
    else:
        clf_pipe = Pipeline(
            [('minMaxScaler', MinMaxScaler()),
             ('PCA', PCA),
             ('CLF',clf)]
        )
    
    colors = cycle(['cyan', 'indigo', 'seagreen', 'yellow', 'blue', 'darkorange', 'pink', 'darkred', 'dimgray', 'maroon', 'coral'])
    
    mean_tpr = 0.0
    mean_fpr = np.linspace(0, 1, 100)
    lw = 2
    i = 0
    
    accuracy = []
    #logloss = []
    
    for (train, test), color in zip(cv.split(X, y), colors):
        clf_pipe.fit(X[train],y[train])  # train object
        y_hat = clf_pipe.predict(X[test]) # get test set preditions
        
        probas_ = clf_pipe.fit(X[train], y[train]).predict_proba(X[test])
        
        # Compute ROC curve and area the curve
        fpr, tpr, thresholds = roc_curve(y[test], probas_[:, 1])
        mean_tpr += interp(mean_fpr, fpr, tpr)
        mean_tpr[0] = 0.0
        roc_auc = auc(fpr, tpr)
        
        plt.rcParams['figure.figsize'] = (12, 6)
        
        plt.plot(fpr, tpr, lw=lw, color=color,
                 label='ROC fold %d (area = %0.2f)' % (i, roc_auc))

        i += 1
    
    plot_ROC_curve(X, y, mean_tpr, mean_fpr)         
        #logloss.append(round(l,5)) 
    
    #print("Accuracy Ratings across all iterations: {0}\n\n\
#Average Accuracy: {1}\n\n\
#Log Loss Values across all iterations: {2}\n\n\
#Average Log Loss: {3}\n".format(accuracy, round(sum(accuracy)/len(accuracy),5), logloss,round(sum(logloss)/len(logloss),5)))

    for (train, test), color in zip(cv.split(X, y), colors):
        clf_pipe.fit(X[train],y[train])  # train object
        y_hat = clf_pipe.predict(X[test]) # get test set preditions
        
        a = float(mt.accuracy_score(y[test],y_hat))
        #l = float(mt.log_loss(y[test], y_hat))
        
        accuracy.append(round(a,5)) 
        
        ytestnames = np.where(y[test] ==  0,'NS','SC')

        yhatnames  = np.where(y_hat ==  0,'NS', 'SC')

        #print(set(list(y_hat)))
        print("confusion matrix\n{0}\n".format(pd.crosstab(ytestnames, yhatnames, rownames = ['True'], colnames = ['Predicted'], margins = True)))

            # Plot non-normalized confusion matrix
        plt.figure()
        plot_confusion_matrixBinary(confusion_matrix(y[test], y_hat), 
                              classes   =["NS",  "SC"], 
                              normalize =True,
                              title     ='Confusion matrix, with normalization')
    
    print("Accuracy Ratings across all iterations: {0}\n\n\
        Average Accuracy: {1}\n".format(accuracy, round(sum(accuracy)/len(accuracy),5)))

    return clf_pipe.named_steps['CLF'], accuracy

CPU times: user 5 µs, sys: 0 ns, total: 5 µs
Wall time: 9.54 µs

%%time

rfc_clf = RandomForestClassifier(n_estimators       = 10, 
                                 max_features       = 'auto', 
                                 max_depth          = 17.0, 
                                 min_samples_split  = 48, 
                                 min_samples_leaf   = 24,
                                 n_jobs             = -1, 
                                 random_state       = seed) # get object
    
rfc_clf, rfc_acc = compute_kfold_scores_ClassificationBinary(rfc_clf, 
                                                             ##PCA = PCA(n_components=22, svd_solver='randomized', random_state = seed),
                                                             cols = PCList)

/usr/local/es7/lib/python3.5/site-packages/matplotlib/font_manager.py:1297: UserWarning: findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans
  (prop.get_family(), self.defaultFamily[fontext]))

confusion matrix
Predicted    NS    SC   All
True                       
NS          796   704  1500
SC          371  1114  1485
All        1167  1818  2985

Normalized confusion matrix
[[ 0.53066667  0.46933333]
 [ 0.24983165  0.75016835]]

/usr/local/es7/lib/python3.5/site-packages/matplotlib/font_manager.py:1297: UserWarning: findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans
  (prop.get_family(), self.defaultFamily[fontext]))

confusion matrix
Predicted    NS   SC   All
True                      
NS         1350  150  1500
SC          674  811  1485
All        2024  961  2985

Normalized confusion matrix
[[ 0.9         0.1       ]
 [ 0.45387205  0.54612795]]

/usr/local/es7/lib/python3.5/site-packages/matplotlib/font_manager.py:1297: UserWarning: findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans
  (prop.get_family(), self.defaultFamily[fontext]))

confusion matrix
Predicted   NS    SC   All
True                      
NS         773   726  1499
SC          96  1389  1485
All        869  2115  2984

Normalized confusion matrix
[[ 0.51567712  0.48432288]
 [ 0.06464646  0.93535354]]

/usr/local/es7/lib/python3.5/site-packages/matplotlib/font_manager.py:1297: UserWarning: findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans
  (prop.get_family(), self.defaultFamily[fontext]))

confusion matrix
Predicted    NS    SC   All
True                       
NS          944   555  1499
SC          154  1330  1484
All        1098  1885  2983

Normalized confusion matrix
[[ 0.62975317  0.37024683]
 [ 0.10377358  0.89622642]]

/usr/local/es7/lib/python3.5/site-packages/matplotlib/font_manager.py:1297: UserWarning: findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans
  (prop.get_family(), self.defaultFamily[fontext]))

confusion matrix
Predicted    NS    SC   All
True                       
NS          984   515  1499
SC          249  1235  1484
All        1233  1750  2983

Normalized confusion matrix
[[ 0.65643763  0.34356237]
 [ 0.16778976  0.83221024]]

Build a Binary KNN Model¶

KNN¶

Algorithm

Options include "Ball Tree" and "KD Tree".

Ball Trees are binary trees formed from nodes of multidimensional hyperspheres, or "balls". Node hyperspheres may intersect, but each point is assigned to one according to distance from the hypersphere center.
KD Trees are binary trees formed from nodes of multidimensional hyperplanes. Every node in the tree is associated with one of the dimensions, with the hyperplane perpendicular to that dimension's axis.

Our findings, were that the Ball Tree algorithm was considerably less efficient to produce results for all 10 iterations in comparison to the KD Tree Algorithm.

Leaf Size

The size for leaf nodes in the KNN Tree.

Number of Neighbors

After 24 iterations of modifying the above parameters, we land on a final winner based on the highest average Accuracy value across all iterations. Average Accuracy values in our 10 test/train iterations ranged from 66.5216 % from the worst parameter inputs of the Ball_Tree Algorith to a value of 69.5528 % in best tuned KNN Classification model fit. We have chosen to utilize the best input for KD tree, although losing an improvement of .0004 % due to the cost(slower runtime of 07 Minutes 25 Seconds through 10 iterations) of fitting the model as Ball Tree. Parameter inputs for the final K Nearest Neighbor Classification model with the KD Tree Algorithm are as follows:

algorithm	leaf_size	n_neighbors
kd_tree	50	150

%%time


def knn_explorBinary_w_PCA(n_neighbors,
               algorithm ,
               leaf_size,
               PCA,
               Data        = OPMAnalysisDataNoFamBinary,
               cv          = cv,
               seed        = seed):
    startTime = datetime.now()
    y = Data["SEP"].values # get the labels we want    
    
    X = Data.drop("SEP", axis=1).as_matrix() 
    
    knn_clf = KNeighborsClassifier(n_neighbors = n_neighbors, algorithm = algorithm, leaf_size = leaf_size, n_jobs=-1) # get object
    
    # setup pipeline to take PCA, then fit a clf model
    clf_pipe = Pipeline(
        [('minMaxScaler', MinMaxScaler()),
         ('PCA', PCA),
         ('CLF',knn_clf)]
    )

    accuracy = cross_val_score(clf_pipe, X, y, cv=cv.split(X, y)) # this also can help with parallelism
    MeanAccuracy =  sum(accuracy)/len(accuracy)
    accuracy = np.append(accuracy, MeanAccuracy)
    endTime = datetime.now()
    TotalTime = endTime - startTime
    accuracy = np.append(accuracy, TotalTime)
    
    #print(TotalTime)
    #print(accuracy)
    
    return accuracy

CPU times: user 13 µs, sys: 7 µs, total: 20 µs
Wall time: 11.9 µs

%%time

def knn_explorBinary(n_neighbors,
               algorithm ,
               leaf_size,
               Data        = OPMAnalysisDataNoFamBinary,
               cols        = PCList,
               cv          = cv,
               seed        = seed):
    startTime = datetime.now()
    y = Data["SEP"].values # get the labels we want    
    
    if ("SEP" in cols):    X = Data[cols].drop("SEP", axis=1).as_matrix() 
    else: X = Data[cols]
    
    knn_clf = KNeighborsClassifier(n_neighbors = n_neighbors, algorithm = algorithm, leaf_size = leaf_size, n_jobs=-1) # get object
    
    # setup pipeline to take PCA, then fit a clf model
    clf_pipe = Pipeline(
        [('minMaxScaler', MinMaxScaler()),
         ('CLF',knn_clf)]
    )

    accuracy = cross_val_score(clf_pipe, X, y, cv=cv.split(X, y)) # this also can help with parallelism
    MeanAccuracy =  sum(accuracy)/len(accuracy)
    accuracy = np.append(accuracy, MeanAccuracy)
    endTime = datetime.now()
    TotalTime = endTime - startTime
    accuracy = np.append(accuracy, TotalTime)
    
    #print(TotalTime)
    #print(accuracy)
    
    return accuracy

CPU times: user 13 µs, sys: 8 µs, total: 21 µs
Wall time: 12.2 µs

%%time
###Full Columns
acclist = [] 

n_neighbors =  [5          , 10         , 15         , 20         , 30         , 40         , 50         , 100        , 150        , 200        , 250        , 150        , 150        , 150        , 150        , 150        , 150        ]
algorithm   =  'kd_tree'
leaf_size   =  [30         , 30         , 30         , 30         , 30         , 30         , 30         , 30         , 30         , 30         , 30         , 2          , 3          , 4          , 5          , 10         , 20         ]



for i in range(0,len(n_neighbors)):
    acclist.append(knn_explorBinary(n_neighbors = n_neighbors[i],
                              algorithm   = algorithm,
                              leaf_size   = leaf_size[i],
                              cols = fullColumns
                             )
                  )

rfcdf = pd.DataFrame(pd.concat([pd.DataFrame({
                                                "ModelVersion": "KNN: " + algorithm + ", Full Raw Columns",
                                                "n_neighbors": n_neighbors,          
                                                "algorithm": algorithm,         
                                                "leaf_size": leaf_size  
                                              }),
                               pd.DataFrame(acclist)], axis = 1).reindex())
rfcdf.columns = ['ModelVersion','algorithm', 'leaf_size','n_neighbors', 'Iteration 0', 'Iteration 1', 'Iteration 2', 'Iteration 3', 'Iteration 4', 'MeanAccuracy', 'RunTime']
TopResultsDF = pd.concat([TopResultsDF, rfcdf.sort_values(['MeanAccuracy'], ascending=False)[TopResultsDF.columns].head(1)]).sort_values(['MeanAccuracy'], ascending=False).reset_index(drop=True)
display(rfcdf)
del rfcdf, acclist



acclist = [] 

n_neighbors =  [5          , 10         , 15         , 20         , 30         , 40         , 50         , 100        , 150        , 200        , 250        , 150        , 150        , 150        , 150        , 150        , 150        ]
algorithm   =  'ball_tree'
leaf_size   =  [30         , 30         , 30         , 30         , 30         , 30         , 30         , 30         , 30         , 30         , 30         , 2          , 3          , 4          , 5          , 10         , 20         ]

for i in range(0,len(n_neighbors)):
    acclist.append(knn_explorBinary(n_neighbors = n_neighbors[i],
                              algorithm   = algorithm,
                              leaf_size   = leaf_size[i],
                              cols = fullColumns
                             )
                  )

rfcdf = pd.DataFrame(pd.concat([pd.DataFrame({
                                                "ModelVersion": "KNN: " + algorithm + ", Full Raw Columns",
                                                "n_neighbors": n_neighbors,          
                                                "algorithm": algorithm,         
                                                "leaf_size": leaf_size  
                                              }),
                               pd.DataFrame(acclist)], axis = 1).reindex())
rfcdf.columns = ['ModelVersion','algorithm', 'leaf_size','n_neighbors', 'Iteration 0', 'Iteration 1', 'Iteration 2', 'Iteration 3', 'Iteration 4', 'MeanAccuracy', 'RunTime']
display(rfcdf)
TopResultsDF = pd.concat([TopResultsDF, rfcdf.sort_values(['MeanAccuracy'], ascending=False)[TopResultsDF.columns].head(1)]).sort_values(['MeanAccuracy'], ascending=False).reset_index(drop=True)
del rfcdf, acclist

###Reduced Columns

acclist = [] 

n_neighbors =  [5          , 10         , 15         , 20         , 30         , 40         , 50         , 100        , 150        , 200        , 250        , 300        , 350        , 400        ,  50        ,  50        ,  50        ,  50        ,  50        ,  50        ]
algorithm   =  'kd_tree'
leaf_size   =  [30         , 30         , 30         , 30         , 30         , 30         , 30         , 30         , 30         , 30         , 30         , 30         , 30         , 30         , 2          , 3          , 4          , 5          , 10         , 20         ]



for i in range(0,len(n_neighbors)):
    acclist.append(knn_explorBinary(n_neighbors = n_neighbors[i],
                              algorithm   = algorithm,
                              leaf_size   = leaf_size[i]
                             )
                  )

rfcdf = pd.DataFrame(pd.concat([pd.DataFrame({
                                                "ModelVersion": "KNN: " + algorithm + ", Reduced Raw Columns",
                                                "n_neighbors": n_neighbors,          
                                                "algorithm": algorithm,         
                                                "leaf_size": leaf_size  
                                              }),
                               pd.DataFrame(acclist)], axis = 1).reindex())
rfcdf.columns = ['ModelVersion','algorithm', 'leaf_size','n_neighbors', 'Iteration 0', 'Iteration 1', 'Iteration 2', 'Iteration 3', 'Iteration 4', 'MeanAccuracy', 'RunTime']
display(rfcdf)
TopResultsDF = pd.concat([TopResultsDF, rfcdf.sort_values(['MeanAccuracy'], ascending=False)[TopResultsDF.columns].head(1)]).sort_values(['MeanAccuracy'], ascending=False).reset_index(drop=True)
del rfcdf, acclist



acclist = [] 

n_neighbors =  [5          , 10         , 15         , 20         , 30         , 40         , 50         , 100        , 150        , 200        , 250        , 300        , 350        , 400        ,  50        ,  50        ,  50        ,  50        ,  50        ,  50        ]
algorithm   =  'ball_tree'
leaf_size   =  [30         , 30         , 30         , 30         , 30         , 30         , 30         , 30         , 30         , 30         , 30         , 30         , 30         , 30         , 2          , 3          , 4          , 5          , 10         , 20         ]

for i in range(0,len(n_neighbors)):
    acclist.append(knn_explorBinary(n_neighbors = n_neighbors[i],
                              algorithm   = algorithm,
                              leaf_size   = leaf_size[i]
                             )
                  )

rfcdf = pd.DataFrame(pd.concat([pd.DataFrame({
                                                "ModelVersion": "KNN: " + algorithm + ", Reduced Raw Columns",
                                                "n_neighbors": n_neighbors,          
                                                "algorithm": algorithm,         
                                                "leaf_size": leaf_size  
                                              }),
                               pd.DataFrame(acclist)], axis = 1).reindex())
rfcdf.columns = ['ModelVersion','algorithm', 'leaf_size','n_neighbors', 'Iteration 0', 'Iteration 1', 'Iteration 2', 'Iteration 3', 'Iteration 4', 'MeanAccuracy', 'RunTime']
display(rfcdf)
TopResultsDF = pd.concat([TopResultsDF, rfcdf.sort_values(['MeanAccuracy'], ascending=False)[TopResultsDF.columns].head(1)]).sort_values(['MeanAccuracy'], ascending=False).reset_index(drop=True)
del rfcdf, acclist


#### WITH PCA

acclist = [] 

n_neighbors =  [5          , 10         , 15         , 20         , 30         , 40         , 50         , 100        , 150        , 200        , 250        , 100        , 100        , 100        , 100        , 100        , 100        ]
algorithm   =  'kd_tree'
leaf_size   =  [30         , 30         , 30         , 30         , 30         , 30         , 30         , 30         , 30         , 30         , 30         , 2          , 3          , 4          , 5          , 10         , 20         ]



for i in range(0,len(n_neighbors)):
    acclist.append(knn_explorBinary_w_PCA(n_neighbors = n_neighbors[i],
                              algorithm   = algorithm,
                              leaf_size   = leaf_size[i],
                              PCA = PCA(n_components=22, svd_solver='randomized', random_state = seed)
                             )
                  )

rfcdf = pd.DataFrame(pd.concat([pd.DataFrame({
                                                "ModelVersion": "KNN: " + algorithm + ", With PCA",
                                                "n_neighbors": n_neighbors,          
                                                "algorithm": algorithm,         
                                                "leaf_size": leaf_size  
                                              }),
                               pd.DataFrame(acclist)], axis = 1).reindex())
rfcdf.columns = ['ModelVersion','algorithm', 'leaf_size','n_neighbors', 'Iteration 0', 'Iteration 1', 'Iteration 2', 'Iteration 3', 'Iteration 4', 'MeanAccuracy', 'RunTime']
display(rfcdf)
TopResultsDF = pd.concat([TopResultsDF, rfcdf.sort_values(['MeanAccuracy'], ascending=False)[TopResultsDF.columns].head(1)]).sort_values(['MeanAccuracy'], ascending=False).reset_index(drop=True)
del rfcdf, acclist



acclist = [] 

n_neighbors =  [5          , 10         , 15         , 20         , 30         , 40         , 50         , 100        , 150        , 200        , 250        , 100        , 100        , 100        , 100        , 100        , 100        ]
algorithm   =  'ball_tree'
leaf_size   =  [30         , 30         , 30         , 30         , 30         , 30         , 30         , 30         , 30         , 30         , 30         , 2          , 3          , 4          , 5          , 10         , 20         ]

for i in range(0,len(n_neighbors)):
    acclist.append(knn_explorBinary_w_PCA(n_neighbors = n_neighbors[i],
                              algorithm   = algorithm,
                              leaf_size   = leaf_size[i],
                              PCA = PCA(n_components=22, svd_solver='randomized', random_state = seed)
                             )
                  )

rfcdf = pd.DataFrame(pd.concat([pd.DataFrame({
                                                "ModelVersion": "KNN: " + algorithm + ", With PCA",
                                                "n_neighbors": n_neighbors,          
                                                "algorithm": algorithm,         
                                                "leaf_size": leaf_size  
                                              }),
                               pd.DataFrame(acclist)], axis = 1).reindex())
rfcdf.columns = ['ModelVersion','algorithm', 'leaf_size','n_neighbors', 'Iteration 0', 'Iteration 1', 'Iteration 2', 'Iteration 3', 'Iteration 4', 'MeanAccuracy', 'RunTime']
display(rfcdf)
TopResultsDF = pd.concat([TopResultsDF, rfcdf.sort_values(['MeanAccuracy'], ascending=False)[TopResultsDF.columns].head(1)]).sort_values(['MeanAccuracy'], ascending=False).reset_index(drop=True)
del rfcdf, acclist

CPU times: user 13h 13s, sys: 12min 50s, total: 13h 13min 3s
Wall time: 20min 7s

display(TopResultsDF)

plot = TopResultsDF[["Iteration 0","Iteration 1","Iteration 2","Iteration 3","Iteration 4"]].transpose().plot.line(title = "Top Results Among Varying Model Feature Inputs",rot=45)
plot.set_xlabel("Iterations")
plot.set_ylabel("Accuracies")
plot.legend(loc='center left', bbox_to_anchor=(1.01, .5))

FinalResultsDF = pd.concat([FinalResultsDF, TopResultsDF.sort_values(['MeanAccuracy'], ascending=False)[TopResultsDF.columns].head(1)]).sort_values(['MeanAccuracy'], ascending=False).reset_index(drop=True)
TopResultsDF = pd.DataFrame(columns= ['ModelVersion', 'Iteration 0', 'Iteration 1', 'Iteration 2', 'Iteration 3', 'Iteration 4', 'MeanAccuracy'])

/usr/local/es7/lib/python3.5/site-packages/matplotlib/font_manager.py:1297: UserWarning: findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans
  (prop.get_family(), self.defaultFamily[fontext]))

%%time

knn_clf = KNeighborsClassifier(n_neighbors = 150, algorithm = 'ball_tree',leaf_size = 30, n_jobs=-1) # get object

knn_clf, knn_acc = compute_kfold_scores_ClassificationBinary(clf         = knn_clf,
                                                             cols        = fullColumns)

/usr/local/es7/lib/python3.5/site-packages/matplotlib/font_manager.py:1297: UserWarning: findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans
  (prop.get_family(), self.defaultFamily[fontext]))

confusion matrix
Predicted    NS    SC   All
True                       
NS         1022   478  1500
SC          431  1054  1485
All        1453  1532  2985

Normalized confusion matrix
[[ 0.68133333  0.31866667]
 [ 0.29023569  0.70976431]]

/usr/local/es7/lib/python3.5/site-packages/matplotlib/font_manager.py:1297: UserWarning: findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans
  (prop.get_family(), self.defaultFamily[fontext]))

confusion matrix
Predicted    NS    SC   All
True                       
NS         1230   270  1500
SC          438  1047  1485
All        1668  1317  2985

Normalized confusion matrix
[[ 0.82        0.18      ]
 [ 0.29494949  0.70505051]]

/usr/local/es7/lib/python3.5/site-packages/matplotlib/font_manager.py:1297: UserWarning: findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans
  (prop.get_family(), self.defaultFamily[fontext]))

confusion matrix
Predicted    NS    SC   All
True                       
NS          835   664  1499
SC          213  1272  1485
All        1048  1936  2984

Normalized confusion matrix
[[ 0.55703803  0.44296197]
 [ 0.14343434  0.85656566]]

/usr/local/es7/lib/python3.5/site-packages/matplotlib/font_manager.py:1297: UserWarning: findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans
  (prop.get_family(), self.defaultFamily[fontext]))

confusion matrix
Predicted    NS    SC   All
True                       
NS         1055   444  1499
SC          326  1158  1484
All        1381  1602  2983

Normalized confusion matrix
[[ 0.70380254  0.29619746]
 [ 0.21967655  0.78032345]]

/usr/local/es7/lib/python3.5/site-packages/matplotlib/font_manager.py:1297: UserWarning: findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans
  (prop.get_family(), self.defaultFamily[fontext]))

confusion matrix
Predicted    NS    SC   All
True                       
NS         1072   427  1499
SC          391  1093  1484
All        1463  1520  2983

Normalized confusion matrix
[[ 0.71514343  0.28485657]
 [ 0.26347709  0.73652291]]

Logistic Regression¶

We have chosen to manipulate the cost variable (C) within our logistic regression analyzing accuracies at {1.0, .01, .05, 5}. This parameter is essentially an inverted regularization strength equal to 1/lambda per scikit-learn class function code (lambda being the actual regularization item). Therefore, the smaller the cost value the stronger the regularization.

mapping = {'NS':0, 'SC':1}
y = OPMAnalysisDataNoFamBinary.replace({'SEP': mapping})
y = y.SEP

%%R -i OPMAnalysisDataNoFamBinary,fullColumns,y
install.packages("car") ## Selection 55
require(car)
str(OPMAnalysisDataNoFamBinary)
print(unlist(fullColumns))
print(paste("# of SEP observations = ", length(y)))
print(paste("SEP type = ", unique(y)))
print(paste("SEP class = ", class(y)))

/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "AGELVL_B". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "AGELVL_C". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "AGELVL_D". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "AGELVL_E". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "AGELVL_F". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "AGELVL_G". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "AGELVL_H". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "AGELVL_I". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "AGELVL_J". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "AGELVL_K". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "LOC_01". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "LOC_02". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "LOC_04". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "LOC_05". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "LOC_06". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "LOC_08". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "LOC_09". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "LOC_10". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "LOC_11". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "LOC_12". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "LOC_13". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "LOC_15". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "LOC_16". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "LOC_17". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "LOC_18". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "LOC_19". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "LOC_20". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "LOC_21". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "LOC_22". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "LOC_23". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "LOC_24". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "LOC_25". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "LOC_26". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "LOC_27". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "LOC_28". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "LOC_29". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "LOC_30". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "LOC_31". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "LOC_32". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "LOC_33". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "LOC_34". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "LOC_35". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "LOC_36". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "LOC_37". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "LOC_38". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "LOC_39". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "LOC_40". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "LOC_41". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "LOC_42". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "LOC_44". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "LOC_45". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "LOC_46". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "LOC_47". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "LOC_48". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "LOC_49". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "LOC_50". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "LOC_51". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "LOC_53". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "LOC_54". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "LOC_55". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "LOC_56". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "TOA_10". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "TOA_15". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "TOA_20". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "TOA_30". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "TOA_32". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "TOA_35". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "TOA_38". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "TOA_40". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "TOA_42". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "TOA_44". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "TOA_45". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "TOA_48". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "LOCTYP_1". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "PPTYP_1". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "PPGROUP_11". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "PPGROUP_12". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "TOATYP_1". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/robjects/pandas2ri.py:61: UserWarning: Error while trying to convert the column "TOATYP_2". Fall back to string conversion. The error is: Cannot convert numpy array of unsigned values -- R does not have unsigned integers.
  (name, str(e)))
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/rinterface/__init__.py:145: RRuntimeWarning: Installing package into ‘/scratch/cse/amfrye/R/x86_64-redhat-linux-gnu-library/3.4’
(as ‘lib’ is unspecified)

  warnings.warn(x, RRuntimeWarning)

Selection: 55

/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/rinterface/__init__.py:145: RRuntimeWarning: trying URL 'https://mirrors.nics.utk.edu/cran/src/contrib/car_2.1-5.tar.gz'

  warnings.warn(x, RRuntimeWarning)
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/rinterface/__init__.py:145: RRuntimeWarning: Content type 'application/x-gzip'
  warnings.warn(x, RRuntimeWarning)
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/rinterface/__init__.py:145: RRuntimeWarning:  length 628590 bytes (613 KB)

  warnings.warn(x, RRuntimeWarning)
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/rinterface/__init__.py:145: RRuntimeWarning: =
  warnings.warn(x, RRuntimeWarning)
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/rinterface/__init__.py:145: RRuntimeWarning: 

  warnings.warn(x, RRuntimeWarning)
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/rinterface/__init__.py:145: RRuntimeWarning: downloaded 613 KB


  warnings.warn(x, RRuntimeWarning)
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/rinterface/__init__.py:145: RRuntimeWarning: 
  warnings.warn(x, RRuntimeWarning)
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/rinterface/__init__.py:145: RRuntimeWarning: The downloaded source packages are in
	‘/tmp/RtmpivmiIF/downloaded_packages’
  warnings.warn(x, RRuntimeWarning)
/users3/cse/amfrye/.local/lib/python3.5/site-packages/rpy2/rinterface/__init__.py:145: RRuntimeWarning: Loading required package: car

  warnings.warn(x, RRuntimeWarning)

--- Please select a CRAN mirror for use in this session ---
Secure CRAN mirrors 

 1: 0-Cloud [https]                   2: Algeria [https]                
 3: Australia (Canberra) [https]      4: Australia (Melbourne 1) [https]
 5: Australia (Melbourne 2) [https]   6: Australia (Perth) [https]      
 7: Austria [https]                   8: Belgium (Ghent) [https]        
 9: Brazil (PR) [https]              10: Brazil (RJ) [https]            
11: Brazil (SP 1) [https]            12: Brazil (SP 2) [https]          
13: Bulgaria [https]                 14: Chile 1 [https]                
15: Chile 2 [https]                  16: China (Guangzhou) [https]      
17: China (Lanzhou) [https]          18: Colombia (Cali) [https]        
19: Czech Republic [https]           20: Denmark [https]                
21: Ecuador (Cuenca) [https]         22: Estonia [https]                
23: France (Lyon 1) [https]          24: France (Lyon 2) [https]        
25: France (Marseille) [https]       26: France (Montpellier) [https]   
27: France (Paris 2) [https]         28: Germany (Göttingen) [https]    
29: Germany (Münster) [https]        30: Greece [https]                 
31: Iceland [https]                  32: Indonesia (Jakarta) [https]    
33: Ireland [https]                  34: Italy (Padua) [https]          
35: Japan (Tokyo) [https]            36: Malaysia [https]               
37: Mexico (Mexico City) [https]     38: Norway [https]                 
39: Philippines [https]              40: Serbia [https]                 
41: Spain (A Coruña) [https]         42: Spain (Madrid) [https]         
43: Sweden [https]                   44: Switzerland [https]            
45: Turkey (Denizli) [https]         46: Turkey (Mersin) [https]        
47: UK (Bristol) [https]             48: UK (Cambridge) [https]         
49: UK (London 1) [https]            50: USA (CA 1) [https]             
51: USA (IA) [https]                 52: USA (KS) [https]               
53: USA (MI 1) [https]               54: USA (OR) [https]               
55: USA (TN) [https]                 56: USA (TX 1) [https]             
57: Vietnam [https]                  58: (other mirrors)                


'data.frame':	14920 obs. of  100 variables:
 $ SEP                          : chr  "NS" "NS" "NS" "NS" ...
 $ GSEGRD                       : num  11 12 11 12 13 12 13 13 11 11 ...
 $ IndAvgSalary                 : num  65898 81219 65898 82168 121939 ...
 $ SalaryOverUnderIndAvg        : num  -4041 -9406 -2807 -6547 -26020 ...
 $ LowerLimitAge                : num  20 25 25 25 25 25 25 25 25 25 ...
 $ YearsToRetirement            : num  37 32 32 32 32 32 32 32 32 32 ...
 $ BLS_FEDERAL_OtherSep_Rate    : num  0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 ...
 $ BLS_FEDERAL_Quits_Rate       : num  0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 ...
 $ BLS_FEDERAL_TotalSep_Level   : int  34 34 34 34 34 34 34 34 34 34 ...
 $ BLS_FEDERAL_JobOpenings_Rate : num  2.1 2.1 2.1 2.1 2.1 2.1 2.1 2.1 2.1 2.1 ...
 $ BLS_FEDERAL_OtherSep_Level   : int  10 10 10 10 10 10 10 10 10 10 ...
 $ BLS_FEDERAL_Quits_Level      : int  11 11 11 11 11 11 11 11 11 11 ...
 $ BLS_FEDERAL_JobOpenings_Level: int  58 58 58 58 58 58 58 58 58 58 ...
 $ BLS_FEDERAL_Layoffs_Rate     : num  0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 ...
 $ BLS_FEDERAL_Layoffs_Level    : int  13 13 13 13 13 13 13 13 13 13 ...
 $ BLS_FEDERAL_TotalSep_Rate    : num  1.2 1.2 1.2 1.2 1.2 1.2 1.2 1.2 1.2 1.2 ...
 $ SALARYLog                    : num  11 11.2 11.1 11.2 11.5 ...
 $ LOSSqrt                      : num  2.17 2.68 2 2.41 1.41 ...
 $ SEPCount_EFDATE_OCCLog       : num  5.82 5.82 5.82 4.32 4.14 ...
 $ SEPCount_EFDATE_LOCLog       : num  6.15 6.24 6.83 6.83 6.83 ...
 $ IndAvgSalaryLog              : num  11.1 11.3 11.1 11.3 11.7 ...
 $ AGELVL_B                     : chr  "1" "0" "0" "0" ...
 $ AGELVL_C                     : chr  "0" "1" "1" "1" ...
 $ AGELVL_D                     : chr  "0" "0" "0" "0" ...
 $ AGELVL_E                     : chr  "0" "0" "0" "0" ...
 $ AGELVL_F                     : chr  "0" "0" "0" "0" ...
 $ AGELVL_G                     : chr  "0" "0" "0" "0" ...
 $ AGELVL_H                     : chr  "0" "0" "0" "0" ...
 $ AGELVL_I                     : chr  "0" "0" "0" "0" ...
 $ AGELVL_J                     : chr  "0" "0" "0" "0" ...
 $ AGELVL_K                     : chr  "0" "0" "0" "0" ...
 $ LOC_01                       : chr  "0" "0" "0" "0" ...
 $ LOC_02                       : chr  "0" "0" "0" "0" ...
 $ LOC_04                       : chr  "0" "0" "0" "0" ...
 $ LOC_05                       : chr  "0" "0" "0" "0" ...
 $ LOC_06                       : chr  "0" "0" "0" "0" ...
 $ LOC_08                       : chr  "0" "0" "0" "0" ...
 $ LOC_09                       : chr  "0" "0" "0" "0" ...
 $ LOC_10                       : chr  "0" "0" "0" "0" ...
 $ LOC_11                       : chr  "0" "0" "0" "0" ...
 $ LOC_12                       : chr  "0" "0" "0" "0" ...
 $ LOC_13                       : chr  "0" "0" "0" "0" ...
 $ LOC_15                       : chr  "0" "0" "0" "0" ...
 $ LOC_16                       : chr  "0" "0" "0" "0" ...
 $ LOC_17                       : chr  "0" "0" "0" "0" ...
 $ LOC_18                       : chr  "0" "0" "0" "0" ...
 $ LOC_19                       : chr  "0" "0" "0" "0" ...
 $ LOC_20                       : chr  "0" "0" "0" "0" ...
 $ LOC_21                       : chr  "0" "0" "0" "0" ...
 $ LOC_22                       : chr  "0" "0" "0" "0" ...
 $ LOC_23                       : chr  "0" "0" "0" "0" ...
 $ LOC_24                       : chr  "0" "0" "1" "1" ...
 $ LOC_25                       : chr  "0" "0" "0" "0" ...
 $ LOC_26                       : chr  "0" "0" "0" "0" ...
 $ LOC_27                       : chr  "0" "0" "0" "0" ...
 $ LOC_28                       : chr  "0" "0" "0" "0" ...
 $ LOC_29                       : chr  "0" "0" "0" "0" ...
 $ LOC_30                       : chr  "0" "0" "0" "0" ...
 $ LOC_31                       : chr  "0" "0" "0" "0" ...
 $ LOC_32                       : chr  "0" "0" "0" "0" ...
 $ LOC_33                       : chr  "0" "0" "0" "0" ...
 $ LOC_34                       : chr  "0" "0" "0" "0" ...
 $ LOC_35                       : chr  "0" "0" "0" "0" ...
 $ LOC_36                       : chr  "0" "0" "0" "0" ...
 $ LOC_37                       : chr  "0" "0" "0" "0" ...
 $ LOC_38                       : chr  "0" "0" "0" "0" ...
 $ LOC_39                       : chr  "0" "0" "0" "0" ...
 $ LOC_40                       : chr  "0" "0" "0" "0" ...
 $ LOC_41                       : chr  "0" "0" "0" "0" ...
 $ LOC_42                       : chr  "1" "0" "0" "0" ...
 $ LOC_44                       : chr  "0" "0" "0" "0" ...
 $ LOC_45                       : chr  "0" "0" "0" "0" ...
 $ LOC_46                       : chr  "0" "0" "0" "0" ...
 $ LOC_47                       : chr  "0" "0" "0" "0" ...
 $ LOC_48                       : chr  "0" "0" "0" "0" ...
 $ LOC_49                       : chr  "0" "1" "0" "0" ...
 $ LOC_50                       : chr  "0" "0" "0" "0" ...
 $ LOC_51                       : chr  "0" "0" "0" "0" ...
 $ LOC_53                       : chr  "0" "0" "0" "0" ...
 $ LOC_54                       : chr  "0" "0" "0" "0" ...
 $ LOC_55                       : chr  "0" "0" "0" "0" ...
 $ LOC_56                       : chr  "0" "0" "0" "0" ...
 $ TOA_10                       : chr  "1" "1" "0" "1" ...
 $ TOA_15                       : chr  "0" "0" "0" "0" ...
 $ TOA_20                       : chr  "0" "0" "0" "0" ...
 $ TOA_30                       : chr  "0" "0" "0" "0" ...
 $ TOA_32                       : chr  "0" "0" "0" "0" ...
 $ TOA_35                       : chr  "0" "0" "0" "0" ...
 $ TOA_38                       : chr  "0" "0" "1" "0" ...
 $ TOA_40                       : chr  "0" "0" "0" "0" ...
 $ TOA_42                       : chr  "0" "0" "0" "0" ...
 $ TOA_44                       : chr  "0" "0" "0" "0" ...
 $ TOA_45                       : chr  "0" "0" "0" "0" ...
 $ TOA_48                       : chr  "0" "0" "0" "0" ...
 $ LOCTYP_1                     : chr  "1" "1" "1" "1" ...
 $ PPTYP_1                      : chr  "1" "1" "1" "1" ...
 $ PPGROUP_11                   : chr  "1" "1" "1" "1" ...
 $ PPGROUP_12                   : chr  "0" "0" "0" "0" ...
 $ TOATYP_1                     : chr  "1" "1" "1" "1" ...
  [list output truncated]
 [1] "GSEGRD"                        "IndAvgSalary"                 
 [3] "SalaryOverUnderIndAvg"         "LowerLimitAge"                
 [5] "YearsToRetirement"             "BLS_FEDERAL_OtherSep_Rate"    
 [7] "BLS_FEDERAL_Quits_Rate"        "BLS_FEDERAL_TotalSep_Level"   
 [9] "BLS_FEDERAL_JobOpenings_Rate"  "BLS_FEDERAL_OtherSep_Level"   
[11] "BLS_FEDERAL_Quits_Level"       "BLS_FEDERAL_JobOpenings_Level"
[13] "BLS_FEDERAL_Layoffs_Rate"      "BLS_FEDERAL_Layoffs_Level"    
[15] "BLS_FEDERAL_TotalSep_Rate"     "SALARYLog"                    
[17] "LOSSqrt"                       "SEPCount_EFDATE_OCCLog"       
[19] "SEPCount_EFDATE_LOCLog"        "IndAvgSalaryLog"              
[21] "AGELVL_B"                      "AGELVL_C"                     
[23] "AGELVL_D"                      "AGELVL_E"                     
[25] "AGELVL_F"                      "AGELVL_G"                     
[27] "AGELVL_H"                      "AGELVL_I"                     
[29] "AGELVL_J"                      "AGELVL_K"                     
[31] "LOC_01"                        "LOC_02"                       
[33] "LOC_04"                        "LOC_05"                       
[35] "LOC_06"                        "LOC_08"                       
[37] "LOC_09"                        "LOC_10"                       
[39] "LOC_11"                        "LOC_12"                       
[41] "LOC_13"                        "LOC_15"                       
[43] "LOC_16"                        "LOC_17"                       
[45] "LOC_18"                        "LOC_19"                       
[47] "LOC_20"                        "LOC_21"                       
[49] "LOC_22"                        "LOC_23"                       
[51] "LOC_24"                        "LOC_25"                       
[53] "LOC_26"                        "LOC_27"                       
[55] "LOC_28"                        "LOC_29"                       
[57] "LOC_30"                        "LOC_31"                       
[59] "LOC_32"                        "LOC_33"                       
[61] "LOC_34"                        "LOC_35"                       
[63] "LOC_36"                        "LOC_37"                       
[65] "LOC_38"                        "LOC_39"                       
[67] "LOC_40"                        "LOC_41"                       
[69] "LOC_42"                        "LOC_44"                       
[71] "LOC_45"                        "LOC_46"                       
[73] "LOC_47"                        "LOC_48"                       
[75] "LOC_49"                        "LOC_50"                       
[77] "LOC_51"                        "LOC_53"                       
[79] "LOC_54"                        "LOC_55"                       
[81] "LOC_56"                        "TOA_10"                       
[83] "TOA_15"                        "TOA_20"                       
[85] "TOA_30"                        "TOA_32"                       
[87] "TOA_35"                        "TOA_38"                       
[89] "TOA_40"                        "TOA_42"                       
[91] "TOA_44"                        "TOA_45"                       
[93] "TOA_48"                        "LOCTYP_1"                     
[95] "PPTYP_1"                       "PPGROUP_11"                   
[97] "PPGROUP_12"                    "TOATYP_1"                     
[99] "TOATYP_2"                     
[1] "# of SEP observations =  14920"
[1] "SEP type =  0" "SEP type =  1"
[1] "SEP class =  integer"

LOCTYP_1 and PPTYP_1 have only single level and need removed:

%R sapply(OPMAnalysisDataNoFamBinary, function(x) length(unique(x[!is.na(x)])))

array([    2,     9,  1878, 10766,    10,    10,     4,     4,     9,
          10,     8,     6,    10,     4,     7,     6,  3537,   466,
         180,   399,  1878,     2,     2,     2,     2,     2,     2,
           2,     2,     2,     2,     2,     2,     2,     2,     2,
           2,     2,     2,     2,     2,     2,     2,     2,     2,
           2,     2,     2,     2,     2,     2,     2,     2,     2,
           2,     2,     2,     2,     2,     2,     2,     2,     2,
           2,     2,     2,     2,     2,     2,     2,     2,     2,
           2,     2,     2,     2,     2,     2,     2,     2,     2,
           2,     2,     2,     2,     2,     2,     2,     2,     2,
           2,     2,     2,     2,     1,     1,     2,     2,     2,     2], dtype=int32)

Unique counts with LOCTYP_1 and PPTYP_1 removed:

%R sapply(OPMAnalysisDataNoFamBinary[,-c(95,96)], function(x) length(unique(x[!is.na(x)])))

array([    2,     9,  1878, 10766,    10,    10,     4,     4,     9,
          10,     8,     6,    10,     4,     7,     6,  3537,   466,
         180,   399,  1878,     2,     2,     2,     2,     2,     2,
           2,     2,     2,     2,     2,     2,     2,     2,     2,
           2,     2,     2,     2,     2,     2,     2,     2,     2,
           2,     2,     2,     2,     2,     2,     2,     2,     2,
           2,     2,     2,     2,     2,     2,     2,     2,     2,
           2,     2,     2,     2,     2,     2,     2,     2,     2,
           2,     2,     2,     2,     2,     2,     2,     2,     2,
           2,     2,     2,     2,     2,     2,     2,     2,     2,
           2,     2,     2,     2,     2,     2,     2,     2], dtype=int32)

Data example with LOCTYP_1 and PPTYP_1 removed:

%R OPMAnalysisDataNoFamBinary[,-c(95,96)]

Perform logistic regression with intent to extract only most important features

%%R

vars <- unlist(fullColumns)
vars <- vars[-c(94,95)]
#print(vars)
fla <- paste("SEP ~", paste(vars, collapse="+"))
fla <- as.formula(fla)

OPMAnalysisDataNoFamBinary$SEP <- as.vector(y)
BinLogit <- glm(fla, data = OPMAnalysisDataNoFamBinary, family = "binomial")
summary(BinLogit)

Call:
glm(formula = fla, family = "binomial", data = OPMAnalysisDataNoFamBinary)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-3.0182  -0.6661  -0.1038   0.7300   3.3486  

Coefficients: (8 not defined because of singularities)
                                Estimate Std. Error z value Pr(>|z|)    
(Intercept)                   -3.273e+01  6.386e+00  -5.125 2.97e-07 ***
GSEGRD                        -5.811e-01  4.736e-02 -12.270  < 2e-16 ***
IndAvgSalary                  -1.229e-05  4.672e-06  -2.630 0.008545 ** 
SalaryOverUnderIndAvg         -1.798e-06  6.633e-06  -0.271 0.786282    
LowerLimitAge                 -5.646e-02  2.694e-02  -2.096 0.036085 *  
YearsToRetirement                     NA         NA      NA       NA    
BLS_FEDERAL_OtherSep_Rate      5.803e+01  6.254e+00   9.279  < 2e-16 ***
BLS_FEDERAL_Quits_Rate         1.089e+02  9.746e+00  11.178  < 2e-16 ***
BLS_FEDERAL_TotalSep_Level    -3.277e-01  2.734e-01  -1.198 0.230735    
BLS_FEDERAL_JobOpenings_Rate  -2.227e+01  5.998e+00  -3.713 0.000205 ***
BLS_FEDERAL_OtherSep_Level    -1.462e+00  1.327e-01 -11.023  < 2e-16 ***
BLS_FEDERAL_Quits_Level       -2.072e+00  3.386e-01  -6.118 9.49e-10 ***
BLS_FEDERAL_JobOpenings_Level  7.239e-01  2.008e-01   3.605 0.000312 ***
BLS_FEDERAL_Layoffs_Rate      -6.068e+01  4.466e+00 -13.585  < 2e-16 ***
BLS_FEDERAL_Layoffs_Level      3.960e+00  3.955e-01  10.013  < 2e-16 ***
BLS_FEDERAL_TotalSep_Rate     -3.096e+01  3.016e+00 -10.263  < 2e-16 ***
SALARYLog                      3.755e-01  7.039e-01   0.533 0.593689    
LOSSqrt                       -7.742e-01  2.430e-02 -31.863  < 2e-16 ***
SEPCount_EFDATE_OCCLog        -5.877e-02  1.697e-02  -3.463 0.000534 ***
SEPCount_EFDATE_LOCLog        -4.532e-01  1.305e-01  -3.474 0.000513 ***
IndAvgSalaryLog                4.161e+00  8.560e-01   4.861 1.17e-06 ***
AGELVL_B1                     -9.866e-01  1.154e+00  -0.855 0.392526    
AGELVL_C1                     -3.741e-01  9.996e-01  -0.374 0.708224    
AGELVL_D1                     -1.535e-01  8.643e-01  -0.178 0.859011    
AGELVL_E1                      1.699e-01  7.306e-01   0.233 0.816125    
AGELVL_F1                      4.260e-01  5.976e-01   0.713 0.475969    
AGELVL_G1                      7.197e-01  4.653e-01   1.547 0.121961    
AGELVL_H1                      8.788e-01  3.344e-01   2.628 0.008586 ** 
AGELVL_I1                      7.511e-01  2.081e-01   3.609 0.000307 ***
AGELVL_J1                             NA         NA      NA       NA    
AGELVL_K1                             NA         NA      NA       NA    
LOC_011                        4.648e-02  4.144e-01   0.112 0.910704    
LOC_021                        1.706e-01  4.498e-01   0.379 0.704477    
LOC_041                        8.845e-01  4.196e-01   2.108 0.035044 *  
LOC_051                        1.275e-01  4.344e-01   0.293 0.769216    
LOC_061                        1.442e+00  4.987e-01   2.891 0.003839 ** 
LOC_081                        6.195e-01  4.240e-01   1.461 0.144057    
LOC_091                       -4.799e-01  4.989e-01  -0.962 0.336083    
LOC_101                       -5.719e-01  7.195e-01  -0.795 0.426650    
LOC_111                        8.543e-01  4.826e-01   1.770 0.076687 .  
LOC_121                        7.306e-01  4.375e-01   1.670 0.094945 .  
LOC_131                        7.676e-01  4.281e-01   1.793 0.072989 .  
LOC_151                       -1.256e-02  4.057e-01  -0.031 0.975296    
LOC_161                        1.041e-01  4.578e-01   0.227 0.820080    
LOC_171                        2.223e-01  4.186e-01   0.531 0.595280    
LOC_181                       -4.071e-02  4.451e-01  -0.091 0.927127    
LOC_191                        2.682e-03  4.944e-01   0.005 0.995671    
LOC_201                        7.133e-01  4.503e-01   1.584 0.113183    
LOC_211                        1.217e-01  4.246e-01   0.287 0.774316    
LOC_221                       -6.172e-02  4.211e-01  -0.147 0.883470    
LOC_231                        3.209e-02  5.059e-01   0.063 0.949416    
LOC_241                        7.031e-01  4.569e-01   1.539 0.123862    
LOC_251                        3.305e-01  4.194e-01   0.788 0.430683    
LOC_261                        1.607e-01  4.227e-01   0.380 0.703828    
LOC_271                        5.177e-01  4.434e-01   1.167 0.243021    
LOC_281                        1.093e-01  4.415e-01   0.248 0.804456    
LOC_291                        2.000e-01  4.369e-01   0.458 0.647037    
LOC_301                        7.146e-01  4.403e-01   1.623 0.104571    
LOC_311                       -1.000e-01  4.964e-01  -0.201 0.840309    
LOC_321                        2.666e-01  4.461e-01   0.598 0.550173    
LOC_331                       -2.684e-01  6.484e-01  -0.414 0.678872    
LOC_341                        1.309e-03  4.192e-01   0.003 0.997509    
LOC_351                        7.471e-01  4.073e-01   1.834 0.066598 .  
LOC_361                        5.629e-01  4.389e-01   1.283 0.199644    
LOC_371                        5.401e-01  4.225e-01   1.278 0.201079    
LOC_381                        3.165e-01  5.209e-01   0.608 0.543487    
LOC_391                        2.746e-01  4.189e-01   0.656 0.512098    
LOC_401                        4.496e-01  4.136e-01   1.087 0.276969    
LOC_411                        4.721e-01  4.301e-01   1.097 0.272440    
LOC_421                        2.419e-01  4.348e-01   0.556 0.577926    
LOC_441                       -3.547e-01  6.673e-01  -0.531 0.595110    
LOC_451                        1.987e-01  4.303e-01   0.462 0.644247    
LOC_461                        5.412e-01  4.460e-01   1.213 0.224988    
LOC_471                        4.890e-02  4.352e-01   0.112 0.910533    
LOC_481                        1.274e+00  4.777e-01   2.667 0.007645 ** 
LOC_491                        2.330e-01  4.340e-01   0.537 0.591254    
LOC_501                       -4.656e-01  7.116e-01  -0.654 0.512947    
LOC_511                        6.192e-01  4.712e-01   1.314 0.188832    
LOC_531                        9.418e-01  4.322e-01   2.179 0.029305 *  
LOC_541                       -9.854e-02  4.546e-01  -0.217 0.828404    
LOC_551                       -2.589e-01  4.460e-01  -0.581 0.561521    
LOC_561                               NA         NA      NA       NA    
TOA_101                       -1.148e+00  2.081e-01  -5.516 3.48e-08 ***
TOA_151                       -1.079e+00  2.131e-01  -5.061 4.16e-07 ***
TOA_201                        1.575e-02  2.801e-01   0.056 0.955171    
TOA_301                       -7.135e-01  2.237e-01  -3.189 0.001426 ** 
TOA_321                       -3.497e-01  7.998e-01  -0.437 0.661924    
TOA_351                       -1.503e+00  3.611e-01  -4.161 3.17e-05 ***
TOA_381                       -7.468e-01  2.136e-01  -3.496 0.000472 ***
TOA_401                       -1.049e+00  2.706e-01  -3.878 0.000105 ***
TOA_421                        6.479e-02  4.898e-01   0.132 0.894753    
TOA_441                        1.318e+00  1.131e+00   1.165 0.243998    
TOA_451                        8.104e+00  1.392e+02   0.058 0.953584    
TOA_481                               NA         NA      NA       NA    
PPGROUP_111                    3.799e-01  1.581e-01   2.403 0.016261 *  
PPGROUP_121                           NA         NA      NA       NA    
TOATYP_11                             NA         NA      NA       NA    
TOATYP_21                             NA         NA      NA       NA    
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 20683  on 14919  degrees of freedom
Residual deviance: 13443  on 14830  degrees of freedom
AIC: 13623

Number of Fisher Scoring iterations: 10

%%R
alias(BinLogit)

Model :
SEP ~ GSEGRD + IndAvgSalary + SalaryOverUnderIndAvg + LowerLimitAge + 
    YearsToRetirement + BLS_FEDERAL_OtherSep_Rate + BLS_FEDERAL_Quits_Rate + 
    BLS_FEDERAL_TotalSep_Level + BLS_FEDERAL_JobOpenings_Rate + 
    BLS_FEDERAL_OtherSep_Level + BLS_FEDERAL_Quits_Level + BLS_FEDERAL_JobOpenings_Level + 
    BLS_FEDERAL_Layoffs_Rate + BLS_FEDERAL_Layoffs_Level + BLS_FEDERAL_TotalSep_Rate + 
    SALARYLog + LOSSqrt + SEPCount_EFDATE_OCCLog + SEPCount_EFDATE_LOCLog + 
    IndAvgSalaryLog + AGELVL_B + AGELVL_C + AGELVL_D + AGELVL_E + 
    AGELVL_F + AGELVL_G + AGELVL_H + AGELVL_I + AGELVL_J + AGELVL_K + 
    LOC_01 + LOC_02 + LOC_04 + LOC_05 + LOC_06 + LOC_08 + LOC_09 + 
    LOC_10 + LOC_11 + LOC_12 + LOC_13 + LOC_15 + LOC_16 + LOC_17 + 
    LOC_18 + LOC_19 + LOC_20 + LOC_21 + LOC_22 + LOC_23 + LOC_24 + 
    LOC_25 + LOC_26 + LOC_27 + LOC_28 + LOC_29 + LOC_30 + LOC_31 + 
    LOC_32 + LOC_33 + LOC_34 + LOC_35 + LOC_36 + LOC_37 + LOC_38 + 
    LOC_39 + LOC_40 + LOC_41 + LOC_42 + LOC_44 + LOC_45 + LOC_46 + 
    LOC_47 + LOC_48 + LOC_49 + LOC_50 + LOC_51 + LOC_53 + LOC_54 + 
    LOC_55 + LOC_56 + TOA_10 + TOA_15 + TOA_20 + TOA_30 + TOA_32 + 
    TOA_35 + TOA_38 + TOA_40 + TOA_42 + TOA_44 + TOA_45 + TOA_48 + 
    PPGROUP_11 + PPGROUP_12 + TOATYP_1 + TOATYP_2

Complete :
                  (Intercept) GSEGRD IndAvgSalary SalaryOverUnderIndAvg
YearsToRetirement   57           0      0            0                 
AGELVL_J1           13           0      0            0                 
AGELVL_K1          -12           0      0            0                 
LOC_561              1           0      0            0                 
TOA_481              1           0      0            0                 
PPGROUP_121          1           0      0            0                 
TOATYP_11            0           0      0            0                 
TOATYP_21            1           0      0            0                 
                  LowerLimitAge BLS_FEDERAL_OtherSep_Rate
YearsToRetirement   -1             0                     
AGELVL_J1         -1/5             0                     
AGELVL_K1          1/5             0                     
LOC_561              0             0                     
TOA_481              0             0                     
PPGROUP_121          0             0                     
TOATYP_11            0             0                     
TOATYP_21            0             0                     
                  BLS_FEDERAL_Quits_Rate BLS_FEDERAL_TotalSep_Level
YearsToRetirement    0                      0                      
AGELVL_J1            0                      0                      
AGELVL_K1            0                      0                      
LOC_561              0                      0                      
TOA_481              0                      0                      
PPGROUP_121          0                      0                      
TOATYP_11            0                      0                      
TOATYP_21            0                      0                      
                  BLS_FEDERAL_JobOpenings_Rate BLS_FEDERAL_OtherSep_Level
YearsToRetirement    0                            0                      
AGELVL_J1            0                            0                      
AGELVL_K1            0                            0                      
LOC_561              0                            0                      
TOA_481              0                            0                      
PPGROUP_121          0                            0                      
TOATYP_11            0                            0                      
TOATYP_21            0                            0                      
                  BLS_FEDERAL_Quits_Level BLS_FEDERAL_JobOpenings_Level
YearsToRetirement    0                       0                         
AGELVL_J1            0                       0                         
AGELVL_K1            0                       0                         
LOC_561              0                       0                         
TOA_481              0                       0                         
PPGROUP_121          0                       0                         
TOATYP_11            0                       0                         
TOATYP_21            0                       0                         
                  BLS_FEDERAL_Layoffs_Rate BLS_FEDERAL_Layoffs_Level
YearsToRetirement    0                        0                     
AGELVL_J1            0                        0                     
AGELVL_K1            0                        0                     
LOC_561              0                        0                     
TOA_481              0                        0                     
PPGROUP_121          0                        0                     
TOATYP_11            0                        0                     
TOATYP_21            0                        0                     
                  BLS_FEDERAL_TotalSep_Rate SALARYLog LOSSqrt
YearsToRetirement    0                         0         0   
AGELVL_J1            0                         0         0   
AGELVL_K1            0                         0         0   
LOC_561              0                         0         0   
TOA_481              0                         0         0   
PPGROUP_121          0                         0         0   
TOATYP_11            0                         0         0   
TOATYP_21            0                         0         0   
                  SEPCount_EFDATE_OCCLog SEPCount_EFDATE_LOCLog IndAvgSalaryLog
YearsToRetirement    0                      0                      0           
AGELVL_J1            0                      0                      0           
AGELVL_K1            0                      0                      0           
LOC_561              0                      0                      0           
TOA_481              0                      0                      0           
PPGROUP_121          0                      0                      0           
TOATYP_11            0                      0                      0           
TOATYP_21            0                      0                      0           
                  AGELVL_B1 AGELVL_C1 AGELVL_D1 AGELVL_E1 AGELVL_F1 AGELVL_G1
YearsToRetirement    0         0         0         0         0         0     
AGELVL_J1           -9        -8        -7        -6        -5        -4     
AGELVL_K1            8         7         6         5         4         3     
LOC_561              0         0         0         0         0         0     
TOA_481              0         0         0         0         0         0     
PPGROUP_121          0         0         0         0         0         0     
TOATYP_11            0         0         0         0         0         0     
TOATYP_21            0         0         0         0         0         0     
                  AGELVL_H1 AGELVL_I1 LOC_011 LOC_021 LOC_041 LOC_051 LOC_061
YearsToRetirement    0         0         0       0       0       0       0   
AGELVL_J1           -3        -2         0       0       0       0       0   
AGELVL_K1            2         1         0       0       0       0       0   
LOC_561              0         0        -1      -1      -1      -1      -1   
TOA_481              0         0         0       0       0       0       0   
PPGROUP_121          0         0         0       0       0       0       0   
TOATYP_11            0         0         0       0       0       0       0   
TOATYP_21            0         0         0       0       0       0       0   
                  LOC_081 LOC_091 LOC_101 LOC_111 LOC_121 LOC_131 LOC_151
YearsToRetirement    0       0       0       0       0       0       0   
AGELVL_J1            0       0       0       0       0       0       0   
AGELVL_K1            0       0       0       0       0       0       0   
LOC_561             -1      -1      -1      -1      -1      -1      -1   
TOA_481              0       0       0       0       0       0       0   
PPGROUP_121          0       0       0       0       0       0       0   
TOATYP_11            0       0       0       0       0       0       0   
TOATYP_21            0       0       0       0       0       0       0   
                  LOC_161 LOC_171 LOC_181 LOC_191 LOC_201 LOC_211 LOC_221
YearsToRetirement    0       0       0       0       0       0       0   
AGELVL_J1            0       0       0       0       0       0       0   
AGELVL_K1            0       0       0       0       0       0       0   
LOC_561             -1      -1      -1      -1      -1      -1      -1   
TOA_481              0       0       0       0       0       0       0   
PPGROUP_121          0       0       0       0       0       0       0   
TOATYP_11            0       0       0       0       0       0       0   
TOATYP_21            0       0       0       0       0       0       0   
                  LOC_231 LOC_241 LOC_251 LOC_261 LOC_271 LOC_281 LOC_291
YearsToRetirement    0       0       0       0       0       0       0   
AGELVL_J1            0       0       0       0       0       0       0   
AGELVL_K1            0       0       0       0       0       0       0   
LOC_561             -1      -1      -1      -1      -1      -1      -1   
TOA_481              0       0       0       0       0       0       0   
PPGROUP_121          0       0       0       0       0       0       0   
TOATYP_11            0       0       0       0       0       0       0   
TOATYP_21            0       0       0       0       0       0       0   
                  LOC_301 LOC_311 LOC_321 LOC_331 LOC_341 LOC_351 LOC_361
YearsToRetirement    0       0       0       0       0       0       0   
AGELVL_J1            0       0       0       0       0       0       0   
AGELVL_K1            0       0       0       0       0       0       0   
LOC_561             -1      -1      -1      -1      -1      -1      -1   
TOA_481              0       0       0       0       0       0       0   
PPGROUP_121          0       0       0       0       0       0       0   
TOATYP_11            0       0       0       0       0       0       0   
TOATYP_21            0       0       0       0       0       0       0   
                  LOC_371 LOC_381 LOC_391 LOC_401 LOC_411 LOC_421 LOC_441
YearsToRetirement    0       0       0       0       0       0       0   
AGELVL_J1            0       0       0       0       0       0       0   
AGELVL_K1            0       0       0       0       0       0       0   
LOC_561             -1      -1      -1      -1      -1      -1      -1   
TOA_481              0       0       0       0       0       0       0   
PPGROUP_121          0       0       0       0       0       0       0   
TOATYP_11            0       0       0       0       0       0       0   
TOATYP_21            0       0       0       0       0       0       0   
                  LOC_451 LOC_461 LOC_471 LOC_481 LOC_491 LOC_501 LOC_511
YearsToRetirement    0       0       0       0       0       0       0   
AGELVL_J1            0       0       0       0       0       0       0   
AGELVL_K1            0       0       0       0       0       0       0   
LOC_561             -1      -1      -1      -1      -1      -1      -1   
TOA_481              0       0       0       0       0       0       0   
PPGROUP_121          0       0       0       0       0       0       0   
TOATYP_11            0       0       0       0       0       0       0   
TOATYP_21            0       0       0       0       0       0       0   
                  LOC_531 LOC_541 LOC_551 TOA_101 TOA_151 TOA_201 TOA_301
YearsToRetirement    0       0       0       0       0       0       0   
AGELVL_J1            0       0       0       0       0       0       0   
AGELVL_K1            0       0       0       0       0       0       0   
LOC_561             -1      -1      -1       0       0       0       0   
TOA_481              0       0       0      -1      -1      -1      -1   
PPGROUP_121          0       0       0       0       0       0       0   
TOATYP_11            0       0       0       1       1       0       1   
TOATYP_21            0       0       0      -1      -1       0      -1   
                  TOA_321 TOA_351 TOA_381 TOA_401 TOA_421 TOA_441 TOA_451
YearsToRetirement    0       0       0       0       0       0       0   
AGELVL_J1            0       0       0       0       0       0       0   
AGELVL_K1            0       0       0       0       0       0       0   
LOC_561              0       0       0       0       0       0       0   
TOA_481             -1      -1      -1      -1      -1      -1      -1   
PPGROUP_121          0       0       0       0       0       0       0   
TOATYP_11            1       1       1       0       0       0       0   
TOATYP_21           -1      -1      -1       0       0       0       0   
                  PPGROUP_111
YearsToRetirement    0       
AGELVL_J1            0       
AGELVL_K1            0       
LOC_561              0       
TOA_481              0       
PPGROUP_121         -1       
TOATYP_11            0       
TOATYP_21            0

%%R
tmp <- alias(BinLogit)$Complete
print(attributes(tmp))
aliased <- dimnames(tmp)[[1]]

$dim
[1]  8 90

$dimnames
$dimnames[[1]]
[1] "YearsToRetirement" "AGELVL_J1"         "AGELVL_K1"        
[4] "LOC_561"           "TOA_481"           "PPGROUP_121"      
[7] "TOATYP_11"         "TOATYP_21"        

$dimnames[[2]]
 [1] "(Intercept)"                   "GSEGRD"                       
 [3] "IndAvgSalary"                  "SalaryOverUnderIndAvg"        
 [5] "LowerLimitAge"                 "BLS_FEDERAL_OtherSep_Rate"    
 [7] "BLS_FEDERAL_Quits_Rate"        "BLS_FEDERAL_TotalSep_Level"   
 [9] "BLS_FEDERAL_JobOpenings_Rate"  "BLS_FEDERAL_OtherSep_Level"   
[11] "BLS_FEDERAL_Quits_Level"       "BLS_FEDERAL_JobOpenings_Level"
[13] "BLS_FEDERAL_Layoffs_Rate"      "BLS_FEDERAL_Layoffs_Level"    
[15] "BLS_FEDERAL_TotalSep_Rate"     "SALARYLog"                    
[17] "LOSSqrt"                       "SEPCount_EFDATE_OCCLog"       
[19] "SEPCount_EFDATE_LOCLog"        "IndAvgSalaryLog"              
[21] "AGELVL_B1"                     "AGELVL_C1"                    
[23] "AGELVL_D1"                     "AGELVL_E1"                    
[25] "AGELVL_F1"                     "AGELVL_G1"                    
[27] "AGELVL_H1"                     "AGELVL_I1"                    
[29] "LOC_011"                       "LOC_021"                      
[31] "LOC_041"                       "LOC_051"                      
[33] "LOC_061"                       "LOC_081"                      
[35] "LOC_091"                       "LOC_101"                      
[37] "LOC_111"                       "LOC_121"                      
[39] "LOC_131"                       "LOC_151"                      
[41] "LOC_161"                       "LOC_171"                      
[43] "LOC_181"                       "LOC_191"                      
[45] "LOC_201"                       "LOC_211"                      
[47] "LOC_221"                       "LOC_231"                      
[49] "LOC_241"                       "LOC_251"                      
[51] "LOC_261"                       "LOC_271"                      
[53] "LOC_281"                       "LOC_291"                      
[55] "LOC_301"                       "LOC_311"                      
[57] "LOC_321"                       "LOC_331"                      
[59] "LOC_341"                       "LOC_351"                      
[61] "LOC_361"                       "LOC_371"                      
[63] "LOC_381"                       "LOC_391"                      
[65] "LOC_401"                       "LOC_411"                      
[67] "LOC_421"                       "LOC_441"                      
[69] "LOC_451"                       "LOC_461"                      
[71] "LOC_471"                       "LOC_481"                      
[73] "LOC_491"                       "LOC_501"                      
[75] "LOC_511"                       "LOC_531"                      
[77] "LOC_541"                       "LOC_551"                      
[79] "TOA_101"                       "TOA_151"                      
[81] "TOA_201"                       "TOA_301"                      
[83] "TOA_321"                       "TOA_351"                      
[85] "TOA_381"                       "TOA_401"                      
[87] "TOA_421"                       "TOA_441"                      
[89] "TOA_451"                       "PPGROUP_111"                  


$fracs
  [1] "57"   "13"   "-12"  "1"    "1"    "1"    "0"    "1"    "0"    "0"   
 [11] "0"    "0"    "0"    "0"    "0"    "0"    "0"    "0"    "0"    "0"   
 [21] "0"    "0"    "0"    "0"    "0"    "0"    "0"    "0"    "0"    "0"   
 [31] "0"    "0"    "-1"   "-1/5" "1/5"  "0"    "0"    "0"    "0"    "0"   
 [41] "0"    "0"    "0"    "0"    "0"    "0"    "0"    "0"    "0"    "0"   
 [51] "0"    "0"    "0"    "0"    "0"    "0"    "0"    "0"    "0"    "0"   
 [61] "0"    "0"    "0"    "0"    "0"    "0"    "0"    "0"    "0"    "0"   
 [71] "0"    "0"    "0"    "0"    "0"    "0"    "0"    "0"    "0"    "0"   
 [81] "0"    "0"    "0"    "0"    "0"    "0"    "0"    "0"    "0"    "0"   
 [91] "0"    "0"    "0"    "0"    "0"    "0"    "0"    "0"    "0"    "0"   
[101] "0"    "0"    "0"    "0"    "0"    "0"    "0"    "0"    "0"    "0"   
[111] "0"    "0"    "0"    "0"    "0"    "0"    "0"    "0"    "0"    "0"   
[121] "0"    "0"    "0"    "0"    "0"    "0"    "0"    "0"    "0"    "0"   
[131] "0"    "0"    "0"    "0"    "0"    "0"    "0"    "0"    "0"    "0"   
[141] "0"    "0"    "0"    "0"    "0"    "0"    "0"    "0"    "0"    "0"   
[151] "0"    "0"    "0"    "0"    "0"    "0"    "0"    "0"    "0"    "0"   
[161] "0"    "-9"   "8"    "0"    "0"    "0"    "0"    "0"    "0"    "-8"  
[171] "7"    "0"    "0"    "0"    "0"    "0"    "0"    "-7"   "6"    "0"   
[181] "0"    "0"    "0"    "0"    "0"    "-6"   "5"    "0"    "0"    "0"   
[191] "0"    "0"    "0"    "-5"   "4"    "0"    "0"    "0"    "0"    "0"   
[201] "0"    "-4"   "3"    "0"    "0"    "0"    "0"    "0"    "0"    "-3"  
[211] "2"    "0"    "0"    "0"    "0"    "0"    "0"    "-2"   "1"    "0"   
[221] "0"    "0"    "0"    "0"    "0"    "0"    "0"    "-1"   "0"    "0"   
[231] "0"    "0"    "0"    "0"    "0"    "-1"   "0"    "0"    "0"    "0"   
[241] "0"    "0"    "0"    "-1"   "0"    "0"    "0"    "0"    "0"    "0"   
[251] "0"    "-1"   "0"    "0"    "0"    "0"    "0"    "0"    "0"    "-1"  
[261] "0"    "0"    "0"    "0"    "0"    "0"    "0"    "-1"   "0"    "0"   
[271] "0"    "0"    "0"    "0"    "0"    "-1"   "0"    "0"    "0"    "0"   
[281] "0"    "0"    "0"    "-1"   "0"    "0"    "0"    "0"    "0"    "0"   
[291] "0"    "-1"   "0"    "0"    "0"    "0"    "0"    "0"    "0"    "-1"  
[301] "0"    "0"    "0"    "0"    "0"    "0"    "0"    "-1"   "0"    "0"   
[311] "0"    "0"    "0"    "0"    "0"    "-1"   "0"    "0"    "0"    "0"   
[321] "0"    "0"    "0"    "-1"   "0"    "0"    "0"    "0"    "0"    "0"   
[331] "0"    "-1"   "0"    "0"    "0"    "0"    "0"    "0"    "0"    "-1"  
[341] "0"    "0"    "0"    "0"    "0"    "0"    "0"    "-1"   "0"    "0"   
[351] "0"    "0"    "0"    "0"    "0"    "-1"   "0"    "0"    "0"    "0"   
[361] "0"    "0"    "0"    "-1"   "0"    "0"    "0"    "0"    "0"    "0"   
[371] "0"    "-1"   "0"    "0"    "0"    "0"    "0"    "0"    "0"    "-1"  
[381] "0"    "0"    "0"    "0"    "0"    "0"    "0"    "-1"   "0"    "0"   
[391] "0"    "0"    "0"    "0"    "0"    "-1"   "0"    "0"    "0"    "0"   
[401] "0"    "0"    "0"    "-1"   "0"    "0"    "0"    "0"    "0"    "0"   
[411] "0"    "-1"   "0"    "0"    "0"    "0"    "0"    "0"    "0"    "-1"  
[421] "0"    "0"    "0"    "0"    "0"    "0"    "0"    "-1"   "0"    "0"   
[431] "0"    "0"    "0"    "0"    "0"    "-1"   "0"    "0"    "0"    "0"   
[441] "0"    "0"    "0"    "-1"   "0"    "0"    "0"    "0"    "0"    "0"   
[451] "0"    "-1"   "0"    "0"    "0"    "0"    "0"    "0"    "0"    "-1"  
[461] "0"    "0"    "0"    "0"    "0"    "0"    "0"    "-1"   "0"    "0"   
[471] "0"    "0"    "0"    "0"    "0"    "-1"   "0"    "0"    "0"    "0"   
[481] "0"    "0"    "0"    "-1"   "0"    "0"    "0"    "0"    "0"    "0"   
[491] "0"    "-1"   "0"    "0"    "0"    "0"    "0"    "0"    "0"    "-1"  
[501] "0"    "0"    "0"    "0"    "0"    "0"    "0"    "-1"   "0"    "0"   
[511] "0"    "0"    "0"    "0"    "0"    "-1"   "0"    "0"    "0"    "0"   
[521] "0"    "0"    "0"    "-1"   "0"    "0"    "0"    "0"    "0"    "0"   
[531] "0"    "-1"   "0"    "0"    "0"    "0"    "0"    "0"    "0"    "-1"  
[541] "0"    "0"    "0"    "0"    "0"    "0"    "0"    "-1"   "0"    "0"   
[551] "0"    "0"    "0"    "0"    "0"    "-1"   "0"    "0"    "0"    "0"   
[561] "0"    "0"    "0"    "-1"   "0"    "0"    "0"    "0"    "0"    "0"   
[571] "0"    "-1"   "0"    "0"    "0"    "0"    "0"    "0"    "0"    "-1"  
[581] "0"    "0"    "0"    "0"    "0"    "0"    "0"    "-1"   "0"    "0"   
[591] "0"    "0"    "0"    "0"    "0"    "-1"   "0"    "0"    "0"    "0"   
[601] "0"    "0"    "0"    "-1"   "0"    "0"    "0"    "0"    "0"    "0"   
[611] "0"    "-1"   "0"    "0"    "0"    "0"    "0"    "0"    "0"    "-1"  
[621] "0"    "0"    "0"    "0"    "0"    "0"    "0"    "0"    "-1"   "0"   
[631] "1"    "-1"   "0"    "0"    "0"    "0"    "-1"   "0"    "1"    "-1"  
[641] "0"    "0"    "0"    "0"    "-1"   "0"    "0"    "0"    "0"    "0"   
[651] "0"    "0"    "-1"   "0"    "1"    "-1"   "0"    "0"    "0"    "0"   
[661] "-1"   "0"    "1"    "-1"   "0"    "0"    "0"    "0"    "-1"   "0"   
[671] "1"    "-1"   "0"    "0"    "0"    "0"    "-1"   "0"    "1"    "-1"  
[681] "0"    "0"    "0"    "0"    "-1"   "0"    "0"    "0"    "0"    "0"   
[691] "0"    "0"    "-1"   "0"    "0"    "0"    "0"    "0"    "0"    "0"   
[701] "-1"   "0"    "0"    "0"    "0"    "0"    "0"    "0"    "-1"   "0"   
[711] "0"    "0"    "0"    "0"    "0"    "0"    "0"    "-1"   "0"    "0"   

$class
[1] "fractions" "matrix"

%%R
aliased <- ifelse(grepl('[[:digit:]]$', aliased), substr(aliased, 1, nchar(aliased)-1), aliased)
print(c("Following attributes will be dropped from model due to multicollinearity:", aliased))

paste(as.character(length(vars) - length(vars[!vars %in% c(aliased)])), "attributes removed from model input")

[1] "Following attributes will be dropped from model due to multicollinearity:"
[2] "YearsToRetirement"                                                        
[3] "AGELVL_J"                                                                 
[4] "AGELVL_K"                                                                 
[5] "LOC_56"                                                                   
[6] "TOA_48"                                                                   
[7] "PPGROUP_12"                                                               
[8] "TOATYP_1"                                                                 
[9] "TOATYP_2"                                                                 
[1] "8 attributes removed from model input"

%%R
runLogit <- function(less, vars){
    fla <- paste("SEP ~", paste(vars, collapse="+"))
    fla <- as.formula(fla)

    binLog <- glm(fla, data = OPMAnalysisDataNoFamBinary, family = "binomial")
    return(binLog)
}

vars <- vars[!(vars %in% c(aliased))]
BinLogit2 <- runLogit(aliased, vars)
summary(BinLogit2)

Call:
glm(formula = fla, family = "binomial", data = OPMAnalysisDataNoFamBinary)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-3.0182  -0.6661  -0.1038   0.7300   3.3486  

Coefficients:
                                Estimate Std. Error z value Pr(>|z|)    
(Intercept)                   -3.273e+01  6.386e+00  -5.125 2.97e-07 ***
GSEGRD                        -5.811e-01  4.736e-02 -12.270  < 2e-16 ***
IndAvgSalary                  -1.229e-05  4.672e-06  -2.630 0.008545 ** 
SalaryOverUnderIndAvg         -1.798e-06  6.633e-06  -0.271 0.786282    
LowerLimitAge                 -5.646e-02  2.694e-02  -2.096 0.036085 *  
BLS_FEDERAL_OtherSep_Rate      5.803e+01  6.254e+00   9.279  < 2e-16 ***
BLS_FEDERAL_Quits_Rate         1.089e+02  9.746e+00  11.178  < 2e-16 ***
BLS_FEDERAL_TotalSep_Level    -3.277e-01  2.734e-01  -1.198 0.230735    
BLS_FEDERAL_JobOpenings_Rate  -2.227e+01  5.998e+00  -3.713 0.000205 ***
BLS_FEDERAL_OtherSep_Level    -1.462e+00  1.327e-01 -11.023  < 2e-16 ***
BLS_FEDERAL_Quits_Level       -2.072e+00  3.386e-01  -6.118 9.49e-10 ***
BLS_FEDERAL_JobOpenings_Level  7.239e-01  2.008e-01   3.605 0.000312 ***
BLS_FEDERAL_Layoffs_Rate      -6.068e+01  4.466e+00 -13.585  < 2e-16 ***
BLS_FEDERAL_Layoffs_Level      3.960e+00  3.955e-01  10.013  < 2e-16 ***
BLS_FEDERAL_TotalSep_Rate     -3.096e+01  3.016e+00 -10.263  < 2e-16 ***
SALARYLog                      3.755e-01  7.039e-01   0.533 0.593689    
LOSSqrt                       -7.742e-01  2.430e-02 -31.863  < 2e-16 ***
SEPCount_EFDATE_OCCLog        -5.877e-02  1.697e-02  -3.463 0.000534 ***
SEPCount_EFDATE_LOCLog        -4.532e-01  1.305e-01  -3.474 0.000513 ***
IndAvgSalaryLog                4.161e+00  8.560e-01   4.861 1.17e-06 ***
AGELVL_B1                     -9.866e-01  1.154e+00  -0.855 0.392526    
AGELVL_C1                     -3.741e-01  9.996e-01  -0.374 0.708224    
AGELVL_D1                     -1.535e-01  8.643e-01  -0.178 0.859011    
AGELVL_E1                      1.699e-01  7.306e-01   0.233 0.816125    
AGELVL_F1                      4.260e-01  5.976e-01   0.713 0.475969    
AGELVL_G1                      7.197e-01  4.653e-01   1.547 0.121961    
AGELVL_H1                      8.788e-01  3.344e-01   2.628 0.008586 ** 
AGELVL_I1                      7.511e-01  2.081e-01   3.609 0.000307 ***
LOC_011                        4.648e-02  4.144e-01   0.112 0.910704    
LOC_021                        1.706e-01  4.498e-01   0.379 0.704477    
LOC_041                        8.845e-01  4.196e-01   2.108 0.035044 *  
LOC_051                        1.275e-01  4.344e-01   0.293 0.769216    
LOC_061                        1.442e+00  4.987e-01   2.891 0.003839 ** 
LOC_081                        6.195e-01  4.240e-01   1.461 0.144057    
LOC_091                       -4.799e-01  4.989e-01  -0.962 0.336083    
LOC_101                       -5.719e-01  7.195e-01  -0.795 0.426650    
LOC_111                        8.543e-01  4.826e-01   1.770 0.076687 .  
LOC_121                        7.306e-01  4.375e-01   1.670 0.094945 .  
LOC_131                        7.676e-01  4.281e-01   1.793 0.072989 .  
LOC_151                       -1.256e-02  4.057e-01  -0.031 0.975296    
LOC_161                        1.041e-01  4.578e-01   0.227 0.820080    
LOC_171                        2.223e-01  4.186e-01   0.531 0.595280    
LOC_181                       -4.071e-02  4.451e-01  -0.091 0.927127    
LOC_191                        2.682e-03  4.944e-01   0.005 0.995671    
LOC_201                        7.133e-01  4.503e-01   1.584 0.113183    
LOC_211                        1.217e-01  4.246e-01   0.287 0.774316    
LOC_221                       -6.172e-02  4.211e-01  -0.147 0.883470    
LOC_231                        3.209e-02  5.059e-01   0.063 0.949416    
LOC_241                        7.031e-01  4.569e-01   1.539 0.123862    
LOC_251                        3.305e-01  4.194e-01   0.788 0.430683    
LOC_261                        1.607e-01  4.227e-01   0.380 0.703828    
LOC_271                        5.177e-01  4.434e-01   1.167 0.243021    
LOC_281                        1.093e-01  4.415e-01   0.248 0.804456    
LOC_291                        2.000e-01  4.369e-01   0.458 0.647037    
LOC_301                        7.146e-01  4.403e-01   1.623 0.104571    
LOC_311                       -1.000e-01  4.964e-01  -0.201 0.840309    
LOC_321                        2.666e-01  4.461e-01   0.598 0.550173    
LOC_331                       -2.684e-01  6.484e-01  -0.414 0.678872    
LOC_341                        1.309e-03  4.192e-01   0.003 0.997509    
LOC_351                        7.471e-01  4.073e-01   1.834 0.066598 .  
LOC_361                        5.629e-01  4.389e-01   1.283 0.199644    
LOC_371                        5.401e-01  4.225e-01   1.278 0.201079    
LOC_381                        3.165e-01  5.209e-01   0.608 0.543487    
LOC_391                        2.746e-01  4.189e-01   0.656 0.512098    
LOC_401                        4.496e-01  4.136e-01   1.087 0.276969    
LOC_411                        4.721e-01  4.301e-01   1.097 0.272440    
LOC_421                        2.419e-01  4.348e-01   0.556 0.577926    
LOC_441                       -3.547e-01  6.673e-01  -0.531 0.595110    
LOC_451                        1.987e-01  4.303e-01   0.462 0.644247    
LOC_461                        5.412e-01  4.460e-01   1.213 0.224988    
LOC_471                        4.890e-02  4.352e-01   0.112 0.910533    
LOC_481                        1.274e+00  4.777e-01   2.667 0.007645 ** 
LOC_491                        2.330e-01  4.340e-01   0.537 0.591254    
LOC_501                       -4.656e-01  7.116e-01  -0.654 0.512947    
LOC_511                        6.192e-01  4.712e-01   1.314 0.188832    
LOC_531                        9.418e-01  4.322e-01   2.179 0.029305 *  
LOC_541                       -9.854e-02  4.546e-01  -0.217 0.828404    
LOC_551                       -2.589e-01  4.460e-01  -0.581 0.561521    
TOA_101                       -1.148e+00  2.081e-01  -5.516 3.48e-08 ***
TOA_151                       -1.079e+00  2.131e-01  -5.061 4.16e-07 ***
TOA_201                        1.575e-02  2.801e-01   0.056 0.955171    
TOA_301                       -7.135e-01  2.237e-01  -3.189 0.001426 ** 
TOA_321                       -3.497e-01  7.998e-01  -0.437 0.661924    
TOA_351                       -1.503e+00  3.611e-01  -4.161 3.17e-05 ***
TOA_381                       -7.468e-01  2.136e-01  -3.496 0.000472 ***
TOA_401                       -1.049e+00  2.706e-01  -3.878 0.000105 ***
TOA_421                        6.479e-02  4.898e-01   0.132 0.894753    
TOA_441                        1.318e+00  1.131e+00   1.165 0.243998    
TOA_451                        8.104e+00  1.392e+02   0.058 0.953584    
PPGROUP_111                    3.799e-01  1.581e-01   2.403 0.016261 *  
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 20683  on 14919  degrees of freedom
Residual deviance: 13443  on 14830  degrees of freedom
AIC: 13623

Number of Fisher Scoring iterations: 10

%%R

#install.packages("car")
#require(car)
library(car)
runVifs <- function(logit){
    tmp <- as.data.frame(vif(logit))
    colnames(tmp) <- "VIF"

    scipen.default <- getOption("scipen")
    options(scipen=999)
    print(tmp)
    options(scipen=scipen.default)
    return(tmp)
}

vifs.BinLogit2 <- runVifs(BinLogit2)

                                       VIF
GSEGRD                           14.110175
IndAvgSalary                     42.900149
SalaryOverUnderIndAvg             9.579701
LowerLimitAge                   219.325838
BLS_FEDERAL_OtherSep_Rate       564.651743
BLS_FEDERAL_Quits_Rate         1052.281028
BLS_FEDERAL_TotalSep_Level    10461.617797
BLS_FEDERAL_JobOpenings_Rate  12261.146763
BLS_FEDERAL_OtherSep_Level      207.403680
BLS_FEDERAL_Quits_Level        1106.678602
BLS_FEDERAL_JobOpenings_Level 11662.910265
BLS_FEDERAL_Layoffs_Rate       1769.294407
BLS_FEDERAL_Layoffs_Level     10430.725186
BLS_FEDERAL_TotalSep_Rate      1755.735030
SALARYLog                       102.628189
LOSSqrt                           1.973772
SEPCount_EFDATE_OCCLog            1.123983
SEPCount_EFDATE_LOCLog           24.988435
IndAvgSalaryLog                 142.077732
AGELVL_B                         27.744404
AGELVL_C                        207.457982
AGELVL_D                        237.573778
AGELVL_E                        149.529939
AGELVL_F                         83.990092
AGELVL_G                         45.565410
AGELVL_H                         23.764489
AGELVL_I                          9.371740
LOC_01                            6.040791
LOC_02                            2.917066
LOC_04                            9.579683
LOC_05                            3.215149
LOC_06                           37.535884
LOC_08                           11.672734
LOC_09                            2.212192
LOC_10                            1.396902
LOC_11                           53.702314
LOC_12                           14.250681
LOC_13                           13.959157
LOC_15                            6.228701
LOC_16                            2.652246
LOC_17                            8.129706
LOC_18                            3.089854
LOC_19                            2.145980
LOC_20                            2.848711
LOC_21                            4.170515
LOC_22                            3.983023
LOC_23                            2.097936
LOC_24                           35.724766
LOC_25                            4.990639
LOC_26                            4.449805
LOC_27                            3.154766
LOC_28                            3.109605
LOC_29                            6.115997
LOC_30                            3.273054
LOC_31                            2.105127
LOC_32                            2.898692
LOC_33                            1.515369
LOC_34                            4.399908
LOC_35                            7.413789
LOC_36                            9.518652
LOC_37                            7.692933
LOC_38                            1.947582
LOC_39                            8.816013
LOC_40                            6.097166
LOC_41                            4.719139
LOC_42                            9.197718
LOC_44                            1.479747
LOC_45                            3.624529
LOC_46                            2.879513
LOC_47                            4.013711
LOC_48                           29.433241
LOC_49                            4.168211
LOC_50                            1.409060
LOC_51                           41.609055
LOC_53                           11.631298
LOC_54                            2.701110
LOC_55                            2.991687
TOA_10                           22.507735
TOA_15                           12.146029
TOA_20                            2.121598
TOA_30                            7.378587
TOA_32                            1.077935
TOA_35                            1.553062
TOA_38                           10.937132
TOA_40                            2.369396
TOA_42                            1.216934
TOA_44                            1.039689
TOA_45                            1.000003
PPGROUP_11                        1.895725

%%R

vifs.BinLogitRepeat <- vifs.BinLogit2
vars.Repeat <- vars

vif.removed <- vector(mode="character", length=0)
ndev.vect <- vector(mode="character", length=0)
ndf.vect <- vector(mode="character", length=0)
pchisq.vect <- vector(mode="character", length=0)
logLik.vect <- vector(mode="character", length=0)
AIC.vect <- vector(mode="character", length=0)
BIC.vect <- vector(mode="character", length=0)


for(i in seq(1,12)){
    remove <- rownames(vifs.BinLogitRepeat)[which.max(vifs.BinLogitRepeat$VIF)]
    vif.removed <- c(vif.removed, remove)
    cat("\n\n\nRemoved BEFORE this step:", remove, "\n")
    vars.Repeat <- vars.Repeat[!(vars.Repeat %in% c(remove))]
    BinLogitRepeat <- runLogit(remove, vars.Repeat)
    print(summary(BinLogitRepeat))
    vifs.BinLogitRepeat <- runVifs(BinLogitRepeat)
    
    ##goodness of fit
    ndev.vect <- c(ndev.vect, with(BinLogitRepeat, null.deviance - deviance))
    ndf.vect <- c(ndf.vect, with(BinLogitRepeat, df.null - df.residual))
    pchisq.vect <- c(pchisq.vect, with(BinLogitRepeat, pchisq(null.deviance - deviance, df.null - df.residual, lower.tail = FALSE)))
    logLik.vect <- c(logLik.vect, logLik(BinLogitRepeat))
    AIC.vect <- c(AIC.vect, AIC(BinLogitRepeat))
    BIC.vect <- c(BIC.vect, BIC(BinLogitRepeat))
}
cat("\nFollowing variables removed based on VIF values (in order of removal):\n")
print(vif.removed)

cat("\n\nNull Deviances (in order):\n")
print(ndev.vect)
cat("\nMin value at iteration = ", which.min(ndev.vect))

cat("\n\nDiff Degrees of Freedom (in order):\n")
print(ndf.vect)
cat("\nMin value at iteration = ", which.min(ndf.vect))

#cat("\n\nP-ChiSquare (in order):\n")
#print(pchisq.vect)
#cat("\nMin value at iteration = ", which.min(pchisq.vect))

cat("\n\nLog Likelihoods (in order):\n")
print(logLik.vect)
cat("\nMin value at iteration = ", which.min(logLik.vect))

cat("\n\nAIC values (in order):\n")
print(AIC.vect)
cat("\nMin value at iteration = ", which.min(AIC.vect))

cat("\n\nBIC values (in order):\n")
print(BIC.vect)
cat("\nMin value at iteration = ", which.min(BIC.vect))



Removed BEFORE this step: BLS_FEDERAL_JobOpenings_Rate 

Call:
glm(formula = fla, family = "binomial", data = OPMAnalysisDataNoFamBinary)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-3.0254  -0.6709  -0.1042   0.7323   3.3514  

Coefficients:
                                Estimate Std. Error z value Pr(>|z|)    
(Intercept)                   -3.682e+01  6.287e+00  -5.857 4.71e-09 ***
GSEGRD                        -5.812e-01  4.733e-02 -12.279  < 2e-16 ***
IndAvgSalary                  -1.243e-05  4.668e-06  -2.662 0.007770 ** 
SalaryOverUnderIndAvg         -1.955e-06  6.634e-06  -0.295 0.768161    
LowerLimitAge                 -5.608e-02  2.693e-02  -2.083 0.037275 *  
BLS_FEDERAL_OtherSep_Rate      3.549e+01  1.457e+00  24.360  < 2e-16 ***
BLS_FEDERAL_Quits_Rate         8.808e+01  7.932e+00  11.105  < 2e-16 ***
BLS_FEDERAL_TotalSep_Level     4.514e-01  1.749e-01   2.581 0.009850 ** 
BLS_FEDERAL_OtherSep_Level    -1.499e+00  1.323e-01 -11.330  < 2e-16 ***
BLS_FEDERAL_Quits_Level       -2.276e+00  3.340e-01  -6.814 9.52e-12 ***
BLS_FEDERAL_JobOpenings_Level -2.126e-02  6.286e-03  -3.382 0.000721 ***
BLS_FEDERAL_Layoffs_Rate      -5.276e+01  3.913e+00 -13.484  < 2e-16 ***
BLS_FEDERAL_Layoffs_Level      2.569e+00  1.237e-01  20.777  < 2e-16 ***
BLS_FEDERAL_TotalSep_Rate     -2.419e+01  2.410e+00 -10.039  < 2e-16 ***
SALARYLog                      3.867e-01  7.038e-01   0.550 0.582662    
LOSSqrt                       -7.741e-01  2.429e-02 -31.871  < 2e-16 ***
SEPCount_EFDATE_OCCLog        -5.978e-02  1.696e-02  -3.525 0.000424 ***
SEPCount_EFDATE_LOCLog        -4.800e-01  1.301e-01  -3.689 0.000225 ***
IndAvgSalaryLog                4.157e+00  8.554e-01   4.860 1.17e-06 ***
AGELVL_B1                     -9.663e-01  1.153e+00  -0.838 0.402062    
AGELVL_C1                     -3.577e-01  9.990e-01  -0.358 0.720271    
AGELVL_D1                     -1.396e-01  8.638e-01  -0.162 0.871624    
AGELVL_E1                      1.813e-01  7.302e-01   0.248 0.803925    
AGELVL_F1                      4.349e-01  5.973e-01   0.728 0.466540    
AGELVL_G1                      7.293e-01  4.650e-01   1.568 0.116798    
AGELVL_H1                      8.853e-01  3.341e-01   2.649 0.008064 ** 
AGELVL_I1                      7.532e-01  2.079e-01   3.622 0.000292 ***
LOC_011                        7.560e-02  4.148e-01   0.182 0.855401    
LOC_021                        1.881e-01  4.500e-01   0.418 0.675928    
LOC_041                        9.319e-01  4.198e-01   2.220 0.026421 *  
LOC_051                        1.276e-01  4.352e-01   0.293 0.769306    
LOC_061                        1.519e+00  4.982e-01   3.048 0.002303 ** 
LOC_081                        6.686e-01  4.242e-01   1.576 0.115006    
LOC_091                       -4.660e-01  4.995e-01  -0.933 0.350843    
LOC_101                       -6.256e-01  7.209e-01  -0.868 0.385444    
LOC_111                        9.291e-01  4.823e-01   1.926 0.054046 .  
LOC_121                        7.799e-01  4.376e-01   1.782 0.074714 .  
LOC_131                        8.112e-01  4.283e-01   1.894 0.058237 .  
LOC_151                        2.002e-02  4.061e-01   0.049 0.960677    
LOC_161                        1.320e-01  4.590e-01   0.288 0.773621    
LOC_171                        2.568e-01  4.188e-01   0.613 0.539833    
LOC_181                       -1.860e-02  4.458e-01  -0.042 0.966724    
LOC_191                       -6.280e-03  4.950e-01  -0.013 0.989877    
LOC_201                        7.418e-01  4.509e-01   1.645 0.099921 .  
LOC_211                        1.609e-01  4.250e-01   0.379 0.704920    
LOC_221                       -4.295e-02  4.216e-01  -0.102 0.918847    
LOC_231                        4.034e-02  5.058e-01   0.080 0.936434    
LOC_241                        7.717e-01  4.567e-01   1.690 0.091071 .  
LOC_251                        3.592e-01  4.198e-01   0.856 0.392177    
LOC_261                        1.758e-01  4.230e-01   0.416 0.677683    
LOC_271                        5.371e-01  4.438e-01   1.210 0.226161    
LOC_281                        1.199e-01  4.420e-01   0.271 0.786223    
LOC_291                        2.495e-01  4.370e-01   0.571 0.568078    
LOC_301                        7.391e-01  4.407e-01   1.677 0.093535 .  
LOC_311                       -7.783e-02  4.964e-01  -0.157 0.875415    
LOC_321                        2.973e-01  4.459e-01   0.667 0.504872    
LOC_331                       -3.139e-01  6.489e-01  -0.484 0.628520    
LOC_341                        1.808e-02  4.196e-01   0.043 0.965627    
LOC_351                        7.770e-01  4.077e-01   1.906 0.056668 .  
LOC_361                        6.124e-01  4.391e-01   1.395 0.163089    
LOC_371                        5.719e-01  4.227e-01   1.353 0.176017    
LOC_381                        3.072e-01  5.212e-01   0.589 0.555670    
LOC_391                        3.126e-01  4.193e-01   0.745 0.455992    
LOC_401                        4.926e-01  4.138e-01   1.190 0.233891    
LOC_411                        5.062e-01  4.304e-01   1.176 0.239583    
LOC_421                        2.969e-01  4.349e-01   0.683 0.494841    
LOC_441                       -3.727e-01  6.658e-01  -0.560 0.575668    
LOC_451                        2.172e-01  4.307e-01   0.504 0.613998    
LOC_461                        5.434e-01  4.462e-01   1.218 0.223253    
LOC_471                        8.097e-02  4.356e-01   0.186 0.852531    
LOC_481                        1.343e+00  4.775e-01   2.812 0.004928 ** 
LOC_491                        2.728e-01  4.337e-01   0.629 0.529372    
LOC_501                       -5.084e-01  7.120e-01  -0.714 0.475178    
LOC_511                        6.877e-01  4.710e-01   1.460 0.144243    
LOC_531                        9.903e-01  4.324e-01   2.290 0.022008 *  
LOC_541                       -9.079e-02  4.551e-01  -0.200 0.841862    
LOC_551                       -2.414e-01  4.466e-01  -0.541 0.588839    
TOA_101                       -1.139e+00  2.079e-01  -5.478 4.31e-08 ***
TOA_151                       -1.068e+00  2.129e-01  -5.017 5.26e-07 ***
TOA_201                        2.039e-02  2.798e-01   0.073 0.941895    
TOA_301                       -6.994e-01  2.234e-01  -3.130 0.001746 ** 
TOA_321                       -3.206e-01  8.046e-01  -0.399 0.690249    
TOA_351                       -1.490e+00  3.609e-01  -4.128 3.65e-05 ***
TOA_381                       -7.377e-01  2.134e-01  -3.457 0.000545 ***
TOA_401                       -1.035e+00  2.705e-01  -3.827 0.000130 ***
TOA_421                        7.756e-02  4.894e-01   0.158 0.874063    
TOA_441                        1.323e+00  1.131e+00   1.170 0.242085    
TOA_451                        8.126e+00  1.393e+02   0.058 0.953469    
PPGROUP_111                    3.712e-01  1.578e-01   2.353 0.018644 *  
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 20683  on 14919  degrees of freedom
Residual deviance: 13457  on 14831  degrees of freedom
AIC: 13635

Number of Fisher Scoring iterations: 10

                                      VIF
GSEGRD                          14.102411
IndAvgSalary                    42.867169
SalaryOverUnderIndAvg            9.596934
LowerLimitAge                  219.397663
BLS_FEDERAL_OtherSep_Rate       30.490313
BLS_FEDERAL_Quits_Rate         698.268418
BLS_FEDERAL_TotalSep_Level    4264.767790
BLS_FEDERAL_OtherSep_Level     204.909502
BLS_FEDERAL_Quits_Level       1081.666363
BLS_FEDERAL_JobOpenings_Level   11.391217
BLS_FEDERAL_Layoffs_Rate      1354.785981
BLS_FEDERAL_Layoffs_Level     1017.591476
BLS_FEDERAL_TotalSep_Rate     1117.131916
SALARYLog                      102.663330
LOSSqrt                          1.974899
SEPCount_EFDATE_OCCLog           1.123968
SEPCount_EFDATE_LOCLog          24.878470
IndAvgSalaryLog                141.984104
AGELVL_B                        27.781513
AGELVL_C                       207.533042
AGELVL_D                       237.591178
AGELVL_E                       149.593271
AGELVL_F                        83.959156
AGELVL_G                        45.563783
AGELVL_H                        23.767317
AGELVL_I                         9.363424
LOC_01                           6.064846
LOC_02                           2.930911
LOC_04                           9.618857
LOC_05                           3.221621
LOC_06                          37.559058
LOC_08                          11.704767
LOC_09                           2.216146
LOC_10                           1.396248
LOC_11                          53.613146
LOC_12                          14.284626
LOC_13                          13.978394
LOC_15                           6.251134
LOC_16                           2.646779
LOC_17                           8.166015
LOC_18                           3.090363
LOC_19                           2.150571
LOC_20                           2.853493
LOC_21                           4.184083
LOC_22                           3.997274
LOC_23                           2.108420
LOC_24                          35.764621
LOC_25                           5.010330
LOC_26                           4.475522
LOC_27                           3.169450
LOC_28                           3.120200
LOC_29                           6.125775
LOC_30                           3.287498
LOC_31                           2.114779
LOC_32                           2.926428
LOC_33                           1.516780
LOC_34                           4.418664
LOC_35                           7.448987
LOC_36                           9.534560
LOC_37                           7.737850
LOC_38                           1.953249
LOC_39                           8.819970
LOC_40                           6.133611
LOC_41                           4.741161
LOC_42                           9.214579
LOC_44                           1.485738
LOC_45                           3.639638
LOC_46                           2.899271
LOC_47                           4.023972
LOC_48                          29.418807
LOC_49                           4.212852
LOC_50                           1.410219
LOC_51                          41.569863
LOC_53                          11.610658
LOC_54                           2.711236
LOC_55                           2.998441
TOA_10                          22.487254
TOA_15                          12.132765
TOA_20                           2.122214
TOA_30                           7.366183
TOA_32                           1.076719
TOA_35                           1.552401
TOA_38                          10.935927
TOA_40                           2.363610
TOA_42                           1.216749
TOA_44                           1.039597
TOA_45                           1.000003
PPGROUP_11                       1.894311


Removed BEFORE this step: BLS_FEDERAL_TotalSep_Level 

Call:
glm(formula = fla, family = "binomial", data = OPMAnalysisDataNoFamBinary)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-3.0328  -0.6692  -0.1043   0.7330   3.3237  

Coefficients:
                                Estimate Std. Error z value Pr(>|z|)    
(Intercept)                   -3.825e+01  6.262e+00  -6.108 1.01e-09 ***
GSEGRD                        -5.803e-01  4.732e-02 -12.262  < 2e-16 ***
IndAvgSalary                  -1.264e-05  4.668e-06  -2.708 0.006768 ** 
SalaryOverUnderIndAvg         -2.024e-06  6.635e-06  -0.305 0.760301    
LowerLimitAge                 -5.575e-02  2.693e-02  -2.070 0.038420 *  
BLS_FEDERAL_OtherSep_Rate      3.611e+01  1.437e+00  25.120  < 2e-16 ***
BLS_FEDERAL_Quits_Rate         7.270e+01  5.219e+00  13.930  < 2e-16 ***
BLS_FEDERAL_OtherSep_Level    -1.197e+00  6.082e-02 -19.678  < 2e-16 ***
BLS_FEDERAL_Quits_Level       -1.509e+00  1.523e-01  -9.908  < 2e-16 ***
BLS_FEDERAL_JobOpenings_Level -1.101e-02  4.884e-03  -2.254 0.024227 *  
BLS_FEDERAL_Layoffs_Rate      -4.466e+01  2.329e+00 -19.178  < 2e-16 ***
BLS_FEDERAL_Layoffs_Level      2.495e+00  1.203e-01  20.749  < 2e-16 ***
BLS_FEDERAL_TotalSep_Rate     -1.846e+01  9.166e-01 -20.133  < 2e-16 ***
SALARYLog                      3.945e-01  7.037e-01   0.561 0.575088    
LOSSqrt                       -7.747e-01  2.429e-02 -31.889  < 2e-16 ***
SEPCount_EFDATE_OCCLog        -5.918e-02  1.696e-02  -3.489 0.000484 ***
SEPCount_EFDATE_LOCLog        -3.755e-01  1.238e-01  -3.032 0.002429 ** 
IndAvgSalaryLog                4.163e+00  8.553e-01   4.867 1.13e-06 ***
AGELVL_B1                     -9.569e-01  1.153e+00  -0.830 0.406660    
AGELVL_C1                     -3.480e-01  9.991e-01  -0.348 0.727615    
AGELVL_D1                     -1.336e-01  8.639e-01  -0.155 0.877130    
AGELVL_E1                      1.852e-01  7.302e-01   0.254 0.799760    
AGELVL_F1                      4.385e-01  5.973e-01   0.734 0.462897    
AGELVL_G1                      7.301e-01  4.650e-01   1.570 0.116409    
AGELVL_H1                      8.831e-01  3.341e-01   2.643 0.008217 ** 
AGELVL_I1                      7.539e-01  2.079e-01   3.626 0.000288 ***
LOC_011                       -2.489e-02  4.133e-01  -0.060 0.951969    
LOC_021                        1.510e-01  4.498e-01   0.336 0.737064    
LOC_041                        8.026e-01  4.170e-01   1.925 0.054264 .  
LOC_051                        1.123e-01  4.351e-01   0.258 0.796271    
LOC_061                        1.251e+00  4.878e-01   2.565 0.010307 *  
LOC_081                        5.257e-01  4.209e-01   1.249 0.211678    
LOC_091                       -4.061e-01  4.990e-01  -0.814 0.415776    
LOC_101                       -5.356e-01  7.189e-01  -0.745 0.456298    
LOC_111                        6.776e-01  4.728e-01   1.433 0.151837    
LOC_121                        6.051e-01  4.327e-01   1.399 0.161963    
LOC_131                        6.405e-01  4.235e-01   1.512 0.130476    
LOC_151                       -4.301e-02  4.056e-01  -0.106 0.915541    
LOC_161                        1.038e-01  4.589e-01   0.226 0.820988    
LOC_171                        1.353e-01  4.165e-01   0.325 0.745356    
LOC_181                       -7.591e-02  4.456e-01  -0.170 0.864729    
LOC_191                        2.355e-02  4.949e-01   0.048 0.962042    
LOC_201                        7.145e-01  4.509e-01   1.585 0.113007    
LOC_211                        9.266e-02  4.243e-01   0.218 0.827113    
LOC_221                       -9.093e-02  4.213e-01  -0.216 0.829126    
LOC_231                        8.372e-02  5.052e-01   0.166 0.868380    
LOC_241                        5.540e-01  4.493e-01   1.233 0.217637    
LOC_251                        2.784e-01  4.189e-01   0.665 0.506310    
LOC_261                        9.719e-02  4.222e-01   0.230 0.817932    
LOC_271                        4.773e-01  4.436e-01   1.076 0.281896    
LOC_281                        8.483e-02  4.417e-01   0.192 0.847713    
LOC_291                        1.100e-01  4.339e-01   0.253 0.799936    
LOC_301                        6.770e-01  4.402e-01   1.538 0.124030    
LOC_311                       -7.151e-02  4.963e-01  -0.144 0.885438    
LOC_321                        2.643e-01  4.459e-01   0.593 0.553388    
LOC_331                       -2.274e-01  6.493e-01  -0.350 0.726196    
LOC_341                       -4.064e-02  4.191e-01  -0.097 0.922736    
LOC_351                        6.872e-01  4.065e-01   1.691 0.090881 .  
LOC_361                        4.416e-01  4.344e-01   1.017 0.309360    
LOC_371                        4.401e-01  4.199e-01   1.048 0.294601    
LOC_381                        3.386e-01  5.221e-01   0.648 0.516685    
LOC_391                        1.852e-01  4.167e-01   0.444 0.656733    
LOC_401                        4.046e-01  4.127e-01   0.980 0.326849    
LOC_411                        4.058e-01  4.288e-01   0.946 0.344025    
LOC_421                        1.391e-01  4.309e-01   0.323 0.746886    
LOC_441                       -2.710e-01  6.641e-01  -0.408 0.683182    
LOC_451                        1.635e-01  4.303e-01   0.380 0.703905    
LOC_461                        5.476e-01  4.460e-01   1.228 0.219567    
LOC_471                       -8.055e-03  4.345e-01  -0.019 0.985210    
LOC_481                        1.100e+00  4.686e-01   2.347 0.018911 *  
LOC_491                        1.832e-01  4.327e-01   0.423 0.672075    
LOC_501                       -4.124e-01  7.077e-01  -0.583 0.560085    
LOC_511                        4.495e-01  4.623e-01   0.972 0.330963    
LOC_531                        8.255e-01  4.280e-01   1.929 0.053792 .  
LOC_541                       -1.036e-01  4.548e-01  -0.228 0.819877    
LOC_551                       -2.887e-01  4.462e-01  -0.647 0.517593    
TOA_101                       -1.140e+00  2.078e-01  -5.483 4.18e-08 ***
TOA_151                       -1.069e+00  2.128e-01  -5.022 5.11e-07 ***
TOA_201                        1.937e-02  2.797e-01   0.069 0.944790    
TOA_301                       -6.997e-01  2.234e-01  -3.131 0.001740 ** 
TOA_321                       -3.406e-01  8.046e-01  -0.423 0.672023    
TOA_351                       -1.487e+00  3.610e-01  -4.120 3.79e-05 ***
TOA_381                       -7.373e-01  2.133e-01  -3.456 0.000548 ***
TOA_401                       -1.028e+00  2.703e-01  -3.802 0.000143 ***
TOA_421                        7.901e-02  4.895e-01   0.161 0.871768    
TOA_441                        1.331e+00  1.131e+00   1.177 0.239227    
TOA_451                        8.162e+00  1.391e+02   0.059 0.953197    
PPGROUP_111                    3.640e-01  1.578e-01   2.306 0.021084 *  
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 20683  on 14919  degrees of freedom
Residual deviance: 13464  on 14832  degrees of freedom
AIC: 13640

Number of Fisher Scoring iterations: 10

                                     VIF
GSEGRD                         14.096450
IndAvgSalary                   42.844549
SalaryOverUnderIndAvg           9.594943
LowerLimitAge                 219.535659
BLS_FEDERAL_OtherSep_Rate      29.584276
BLS_FEDERAL_Quits_Rate        300.982891
BLS_FEDERAL_OtherSep_Level     43.194480
BLS_FEDERAL_Quits_Level       223.370966
BLS_FEDERAL_JobOpenings_Level   6.852760
BLS_FEDERAL_Layoffs_Rate      479.157701
BLS_FEDERAL_Layoffs_Level     961.396403
BLS_FEDERAL_TotalSep_Rate     161.354837
SALARYLog                     102.645636
LOSSqrt                         1.974931
SEPCount_EFDATE_OCCLog          1.123858
SEPCount_EFDATE_LOCLog         22.542581
IndAvgSalaryLog               141.961235
AGELVL_B                       27.854483
AGELVL_C                      207.826584
AGELVL_D                      237.746047
AGELVL_E                      149.661959
AGELVL_F                       84.060896
AGELVL_G                       45.615524
AGELVL_H                       23.780677
AGELVL_I                        9.360065
LOC_01                          6.017956
LOC_02                          2.935738
LOC_04                          9.525681
LOC_05                          3.231022
LOC_06                         36.005770
LOC_08                         11.534754
LOC_09                          2.214122
LOC_10                          1.395777
LOC_11                         51.513697
LOC_12                         13.994128
LOC_13                         13.681734
LOC_15                          6.251670
LOC_16                          2.648036
LOC_17                          8.078119
LOC_18                          3.082752
LOC_19                          2.152426
LOC_20                          2.857458
LOC_21                          4.178398
LOC_22                          3.997988
LOC_23                          2.112516
LOC_24                         34.606682
LOC_25                          4.988039
LOC_26                          4.458787
LOC_27                          3.158467
LOC_28                          3.128964
LOC_29                          6.056461
LOC_30                          3.277764
LOC_31                          2.118724
LOC_32                          2.925017
LOC_33                          1.510859
LOC_34                          4.423038
LOC_35                          7.410529
LOC_36                          9.336726
LOC_37                          7.638325
LOC_38                          1.947152
LOC_39                          8.704343
LOC_40                          6.097745
LOC_41                          4.710050
LOC_42                          9.070910
LOC_44                          1.482675
LOC_45                          3.640281
LOC_46                          2.910038
LOC_47                          4.004226
LOC_48                         28.340768
LOC_49                          4.180681
LOC_50                          1.413411
LOC_51                         40.067925
LOC_53                         11.382561
LOC_54                          2.722130
LOC_55                          3.001285
TOA_10                         22.494780
TOA_15                         12.133793
TOA_20                          2.122048
TOA_30                          7.365433
TOA_32                          1.076605
TOA_35                          1.551310
TOA_38                         10.941032
TOA_40                          2.367088
TOA_42                          1.216410
TOA_44                          1.039596
TOA_45                          1.000003
PPGROUP_11                      1.893480


Removed BEFORE this step: BLS_FEDERAL_Layoffs_Level 

Call:
glm(formula = fla, family = "binomial", data = OPMAnalysisDataNoFamBinary)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-3.0115  -0.7000  -0.1278   0.7851   3.3366  

Coefficients:
                                Estimate Std. Error z value Pr(>|z|)    
(Intercept)                   -4.301e+01  6.156e+00  -6.986 2.83e-12 ***
GSEGRD                        -5.595e-01  4.647e-02 -12.041  < 2e-16 ***
IndAvgSalary                  -1.328e-05  4.595e-06  -2.890 0.003849 ** 
SalaryOverUnderIndAvg         -1.306e-06  6.577e-06  -0.199 0.842590    
LowerLimitAge                 -5.586e-02  2.664e-02  -2.097 0.036012 *  
BLS_FEDERAL_OtherSep_Rate      1.441e+01  9.446e-01  15.251  < 2e-16 ***
BLS_FEDERAL_Quits_Rate        -2.888e+01  1.885e+00 -15.319  < 2e-16 ***
BLS_FEDERAL_OtherSep_Level    -3.046e-01  4.075e-02  -7.475 7.74e-14 ***
BLS_FEDERAL_Quits_Level        1.250e+00  7.716e-02  16.202  < 2e-16 ***
BLS_FEDERAL_JobOpenings_Level  6.307e-02  3.425e-03  18.417  < 2e-16 ***
BLS_FEDERAL_Layoffs_Rate       2.640e+00  4.573e-01   5.773 7.79e-09 ***
BLS_FEDERAL_TotalSep_Rate     -2.309e+00  4.736e-01  -4.876 1.08e-06 ***
SALARYLog                      2.788e-01  6.943e-01   0.402 0.688043    
LOSSqrt                       -7.634e-01  2.390e-02 -31.935  < 2e-16 ***
SEPCount_EFDATE_OCCLog        -4.145e-02  1.658e-02  -2.500 0.012417 *  
SEPCount_EFDATE_LOCLog        -3.382e-01  1.209e-01  -2.798 0.005147 ** 
IndAvgSalaryLog                4.197e+00  8.427e-01   4.981 6.33e-07 ***
AGELVL_B1                     -9.617e-01  1.140e+00  -0.843 0.398987    
AGELVL_C1                     -3.829e-01  9.882e-01  -0.388 0.698373    
AGELVL_D1                     -2.021e-01  8.545e-01  -0.237 0.813033    
AGELVL_E1                      1.062e-01  7.222e-01   0.147 0.883119    
AGELVL_F1                      3.640e-01  5.906e-01   0.616 0.537678    
AGELVL_G1                      6.340e-01  4.597e-01   1.379 0.167900    
AGELVL_H1                      8.107e-01  3.302e-01   2.455 0.014080 *  
AGELVL_I1                      6.997e-01  2.054e-01   3.407 0.000657 ***
LOC_011                       -1.114e-01  4.073e-01  -0.273 0.784470    
LOC_021                        1.251e-01  4.408e-01   0.284 0.776495    
LOC_041                        6.939e-01  4.106e-01   1.690 0.091053 .  
LOC_051                       -3.647e-02  4.272e-01  -0.085 0.931970    
LOC_061                        1.061e+00  4.794e-01   2.213 0.026896 *  
LOC_081                        4.094e-01  4.141e-01   0.989 0.322851    
LOC_091                       -4.186e-01  4.854e-01  -0.862 0.388531    
LOC_101                       -5.427e-01  7.058e-01  -0.769 0.441931    
LOC_111                        5.141e-01  4.647e-01   1.106 0.268631    
LOC_121                        4.190e-01  4.257e-01   0.984 0.324987    
LOC_131                        4.778e-01  4.166e-01   1.147 0.251420    
LOC_151                       -9.066e-02  3.992e-01  -0.227 0.820361    
LOC_161                       -3.610e-02  4.510e-01  -0.080 0.936203    
LOC_171                       -4.924e-02  4.098e-01  -0.120 0.904368    
LOC_181                       -2.057e-01  4.380e-01  -0.470 0.638567    
LOC_191                       -1.762e-02  4.829e-01  -0.036 0.970893    
LOC_201                        5.917e-01  4.412e-01   1.341 0.179854    
LOC_211                        8.412e-03  4.172e-01   0.020 0.983914    
LOC_221                       -1.913e-01  4.142e-01  -0.462 0.644123    
LOC_231                       -3.702e-02  4.980e-01  -0.074 0.940752    
LOC_241                        3.930e-01  4.416e-01   0.890 0.373525    
LOC_251                        2.137e-01  4.108e-01   0.520 0.602867    
LOC_261                        7.013e-03  4.166e-01   0.017 0.986570    
LOC_271                        3.258e-01  4.344e-01   0.750 0.453240    
LOC_281                       -1.006e-01  4.353e-01  -0.231 0.817261    
LOC_291                       -7.823e-02  4.271e-01  -0.183 0.854680    
LOC_301                        5.859e-01  4.354e-01   1.346 0.178428    
LOC_311                       -4.788e-02  4.887e-01  -0.098 0.921945    
LOC_321                        1.512e-01  4.369e-01   0.346 0.729279    
LOC_331                       -1.695e-01  6.453e-01  -0.263 0.792798    
LOC_341                       -1.284e-01  4.115e-01  -0.312 0.755048    
LOC_351                        5.401e-01  4.005e-01   1.349 0.177458    
LOC_361                        3.195e-01  4.272e-01   0.748 0.454550    
LOC_371                        3.255e-01  4.135e-01   0.787 0.431070    
LOC_381                        2.216e-01  5.142e-01   0.431 0.666465    
LOC_391                       -9.349e-03  4.103e-01  -0.023 0.981821    
LOC_401                        3.098e-01  4.056e-01   0.764 0.445001    
LOC_411                        2.935e-01  4.229e-01   0.694 0.487631    
LOC_421                        1.251e-02  4.242e-01   0.029 0.976477    
LOC_441                       -2.957e-01  6.472e-01  -0.457 0.647706    
LOC_451                        2.051e-02  4.243e-01   0.048 0.961442    
LOC_461                        5.812e-01  4.397e-01   1.322 0.186200    
LOC_471                       -1.252e-01  4.282e-01  -0.292 0.770029    
LOC_481                        8.834e-01  4.604e-01   1.919 0.055020 .  
LOC_491                        3.150e-02  4.257e-01   0.074 0.941023    
LOC_501                       -6.199e-01  6.984e-01  -0.888 0.374790    
LOC_511                        3.074e-01  4.546e-01   0.676 0.498843    
LOC_531                        7.209e-01  4.215e-01   1.710 0.087190 .  
LOC_541                       -2.475e-01  4.481e-01  -0.552 0.580680    
LOC_551                       -3.552e-01  4.389e-01  -0.809 0.418429    
TOA_101                       -1.122e+00  2.031e-01  -5.527 3.25e-08 ***
TOA_151                       -1.070e+00  2.082e-01  -5.141 2.73e-07 ***
TOA_201                        4.745e-02  2.746e-01   0.173 0.862841    
TOA_301                       -7.185e-01  2.184e-01  -3.290 0.001003 ** 
TOA_321                       -4.425e-01  8.108e-01  -0.546 0.585256    
TOA_351                       -1.555e+00  3.537e-01  -4.395 1.11e-05 ***
TOA_381                       -7.398e-01  2.086e-01  -3.546 0.000390 ***
TOA_401                       -1.025e+00  2.648e-01  -3.872 0.000108 ***
TOA_421                        1.070e-01  4.846e-01   0.221 0.825281    
TOA_441                        9.625e-01  1.098e+00   0.876 0.380817    
TOA_451                        8.185e+00  1.344e+02   0.061 0.951423    
PPGROUP_111                    2.733e-01  1.548e-01   1.765 0.077573 .  
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 20683  on 14919  degrees of freedom
Residual deviance: 13915  on 14833  degrees of freedom
AIC: 14089

Number of Fisher Scoring iterations: 10

                                     VIF
GSEGRD                         14.084820
IndAvgSalary                   42.946627
SalaryOverUnderIndAvg           9.659062
LowerLimitAge                 222.204956
BLS_FEDERAL_OtherSep_Rate      13.419915
BLS_FEDERAL_Quits_Rate         38.717795
BLS_FEDERAL_OtherSep_Level     20.402730
BLS_FEDERAL_Quits_Level        55.691958
BLS_FEDERAL_JobOpenings_Level   3.613481
BLS_FEDERAL_Layoffs_Rate       18.823252
BLS_FEDERAL_TotalSep_Rate      44.241110
SALARYLog                     103.354860
LOSSqrt                         1.954522
SEPCount_EFDATE_OCCLog          1.120809
SEPCount_EFDATE_LOCLog         22.385361
IndAvgSalaryLog               142.796733
AGELVL_B                       28.604589
AGELVL_C                      211.730999
AGELVL_D                      242.703461
AGELVL_E                      153.427433
AGELVL_F                       86.336131
AGELVL_G                       46.508626
AGELVL_H                       24.243731
AGELVL_I                        9.382183
LOC_01                          5.976635
LOC_02                          3.003151
LOC_04                          9.529767
LOC_05                          3.271838
LOC_06                         35.861969
LOC_08                         11.622783
LOC_09                          2.276671
LOC_10                          1.398664
LOC_11                         51.904733
LOC_12                         14.130208
LOC_13                         13.959043
LOC_15                          6.339311
LOC_16                          2.665564
LOC_17                          8.090489
LOC_18                          3.107284
LOC_19                          2.194706
LOC_20                          2.925662
LOC_21                          4.229631
LOC_22                          4.044529
LOC_23                          2.101939
LOC_24                         35.047823
LOC_25                          5.170394
LOC_26                          4.400034
LOC_27                          3.237243
LOC_28                          3.121063
LOC_29                          6.334707
LOC_30                          3.236488
LOC_31                          2.118286
LOC_32                          2.982528
LOC_33                          1.493842
LOC_34                          4.507625
LOC_35                          7.329880
LOC_36                          9.438289
LOC_37                          7.597933
LOC_38                          1.942086
LOC_39                          8.669860
LOC_40                          6.228079
LOC_41                          4.673255
LOC_42                          8.979397
LOC_44                          1.495281
LOC_45                          3.621144
LOC_46                          2.900952
LOC_47                          3.994933
LOC_48                         28.968764
LOC_49                          4.200175
LOC_50                          1.410371
LOC_51                         40.230985
LOC_53                         11.359629
LOC_54                          2.715109
LOC_55                          3.016934
TOA_10                         22.333791
TOA_15                         12.080290
TOA_20                          2.100723
TOA_30                          7.351089
TOA_32                          1.071734
TOA_35                          1.551303
TOA_38                         10.828279
TOA_40                          2.348290
TOA_42                          1.209936
TOA_44                          1.039635
TOA_45                          1.000003
PPGROUP_11                      1.882840


Removed BEFORE this step: AGELVL_D 

Call:
glm(formula = fla, family = "binomial", data = OPMAnalysisDataNoFamBinary)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-3.0113  -0.7006  -0.1276   0.7849   3.3311  

Coefficients:
                                Estimate Std. Error z value Pr(>|z|)    
(Intercept)                   -4.340e+01  5.931e+00  -7.317 2.53e-13 ***
GSEGRD                        -5.595e-01  4.647e-02 -12.040  < 2e-16 ***
IndAvgSalary                  -1.329e-05  4.595e-06  -2.892 0.003832 ** 
SalaryOverUnderIndAvg         -1.312e-06  6.577e-06  -0.200 0.841859    
LowerLimitAge                 -4.960e-02  2.770e-03 -17.903  < 2e-16 ***
BLS_FEDERAL_OtherSep_Rate      1.440e+01  9.445e-01  15.250  < 2e-16 ***
BLS_FEDERAL_Quits_Rate        -2.888e+01  1.885e+00 -15.322  < 2e-16 ***
BLS_FEDERAL_OtherSep_Level    -3.045e-01  4.075e-02  -7.473 7.85e-14 ***
BLS_FEDERAL_Quits_Level        1.250e+00  7.716e-02  16.205  < 2e-16 ***
BLS_FEDERAL_JobOpenings_Level  6.307e-02  3.425e-03  18.416  < 2e-16 ***
BLS_FEDERAL_Layoffs_Rate       2.641e+00  4.573e-01   5.775 7.68e-09 ***
BLS_FEDERAL_TotalSep_Rate     -2.310e+00  4.736e-01  -4.877 1.08e-06 ***
SALARYLog                      2.798e-01  6.943e-01   0.403 0.686888    
LOSSqrt                       -7.632e-01  2.389e-02 -31.943  < 2e-16 ***
SEPCount_EFDATE_OCCLog        -4.141e-02  1.658e-02  -2.498 0.012490 *  
SEPCount_EFDATE_LOCLog        -3.388e-01  1.209e-01  -2.803 0.005070 ** 
IndAvgSalaryLog                4.196e+00  8.426e-01   4.980 6.35e-07 ***
AGELVL_B1                     -6.976e-01  2.305e-01  -3.027 0.002471 ** 
AGELVL_C1                     -1.502e-01  8.775e-02  -1.711 0.087052 .  
AGELVL_E1                      2.762e-01  6.920e-02   3.991 6.57e-05 ***
AGELVL_F1                      5.027e-01  7.211e-02   6.971 3.14e-12 ***
AGELVL_G1                      7.412e-01  7.599e-02   9.754  < 2e-16 ***
AGELVL_H1                      8.866e-01  7.823e-02  11.333  < 2e-16 ***
AGELVL_I1                      7.442e-01  8.273e-02   8.996  < 2e-16 ***
LOC_011                       -1.115e-01  4.074e-01  -0.274 0.784377    
LOC_021                        1.249e-01  4.408e-01   0.283 0.776921    
LOC_041                        6.938e-01  4.107e-01   1.690 0.091121 .  
LOC_051                       -3.744e-02  4.272e-01  -0.088 0.930172    
LOC_061                        1.061e+00  4.794e-01   2.214 0.026852 *  
LOC_081                        4.100e-01  4.142e-01   0.990 0.322200    
LOC_091                       -4.181e-01  4.855e-01  -0.861 0.389181    
LOC_101                       -5.424e-01  7.058e-01  -0.768 0.442198    
LOC_111                        5.145e-01  4.648e-01   1.107 0.268330    
LOC_121                        4.192e-01  4.258e-01   0.985 0.324844    
LOC_131                        4.780e-01  4.167e-01   1.147 0.251381    
LOC_151                       -9.118e-02  3.993e-01  -0.228 0.819362    
LOC_161                       -3.764e-02  4.509e-01  -0.083 0.933484    
LOC_171                       -4.922e-02  4.099e-01  -0.120 0.904404    
LOC_181                       -2.057e-01  4.381e-01  -0.469 0.638725    
LOC_191                       -1.730e-02  4.830e-01  -0.036 0.971433    
LOC_201                        5.917e-01  4.412e-01   1.341 0.179920    
LOC_211                        8.037e-03  4.173e-01   0.019 0.984633    
LOC_221                       -1.919e-01  4.142e-01  -0.463 0.643121    
LOC_231                       -3.743e-02  4.982e-01  -0.075 0.940107    
LOC_241                        3.934e-01  4.417e-01   0.891 0.373092    
LOC_251                        2.131e-01  4.109e-01   0.519 0.604011    
LOC_261                        7.156e-03  4.167e-01   0.017 0.986298    
LOC_271                        3.251e-01  4.344e-01   0.748 0.454295    
LOC_281                       -1.013e-01  4.354e-01  -0.233 0.816057    
LOC_291                       -7.885e-02  4.272e-01  -0.185 0.853542    
LOC_301                        5.862e-01  4.355e-01   1.346 0.178275    
LOC_311                       -4.808e-02  4.888e-01  -0.098 0.921633    
LOC_321                        1.508e-01  4.370e-01   0.345 0.729965    
LOC_331                       -1.691e-01  6.454e-01  -0.262 0.793292    
LOC_341                       -1.288e-01  4.115e-01  -0.313 0.754263    
LOC_351                        5.406e-01  4.005e-01   1.350 0.177123    
LOC_361                        3.195e-01  4.273e-01   0.748 0.454542    
LOC_371                        3.257e-01  4.135e-01   0.788 0.430859    
LOC_381                        2.203e-01  5.142e-01   0.428 0.668306    
LOC_391                       -9.850e-03  4.103e-01  -0.024 0.980849    
LOC_401                        3.092e-01  4.057e-01   0.762 0.445988    
LOC_411                        2.936e-01  4.230e-01   0.694 0.487575    
LOC_421                        1.256e-02  4.242e-01   0.030 0.976378    
LOC_441                       -2.957e-01  6.472e-01  -0.457 0.647803    
LOC_451                        2.025e-02  4.244e-01   0.048 0.961946    
LOC_461                        5.802e-01  4.397e-01   1.320 0.186996    
LOC_471                       -1.261e-01  4.282e-01  -0.295 0.768343    
LOC_481                        8.839e-01  4.605e-01   1.920 0.054902 .  
LOC_491                        3.067e-02  4.258e-01   0.072 0.942572    
LOC_501                       -6.191e-01  6.988e-01  -0.886 0.375617    
LOC_511                        3.078e-01  4.546e-01   0.677 0.498454    
LOC_531                        7.210e-01  4.216e-01   1.710 0.087205 .  
LOC_541                       -2.478e-01  4.482e-01  -0.553 0.580332    
LOC_551                       -3.562e-01  4.390e-01  -0.811 0.417133    
TOA_101                       -1.122e+00  2.030e-01  -5.526 3.28e-08 ***
TOA_151                       -1.070e+00  2.081e-01  -5.138 2.77e-07 ***
TOA_201                        4.860e-02  2.746e-01   0.177 0.859519    
TOA_301                       -7.180e-01  2.184e-01  -3.288 0.001009 ** 
TOA_321                       -4.426e-01  8.105e-01  -0.546 0.585051    
TOA_351                       -1.554e+00  3.537e-01  -4.393 1.12e-05 ***
TOA_381                       -7.389e-01  2.085e-01  -3.543 0.000395 ***
TOA_401                       -1.025e+00  2.648e-01  -3.869 0.000109 ***
TOA_421                        1.081e-01  4.846e-01   0.223 0.823532    
TOA_441                        9.632e-01  1.098e+00   0.877 0.380492    
TOA_451                        8.186e+00  1.344e+02   0.061 0.951417    
PPGROUP_111                    2.739e-01  1.548e-01   1.769 0.076825 .  
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 20683  on 14919  degrees of freedom
Residual deviance: 13915  on 14834  degrees of freedom
AIC: 14087

Number of Fisher Scoring iterations: 10

                                     VIF
GSEGRD                         14.084657
IndAvgSalary                   42.945520
SalaryOverUnderIndAvg           9.658435
LowerLimitAge                   2.404092
BLS_FEDERAL_OtherSep_Rate      13.418547
BLS_FEDERAL_Quits_Rate         38.714968
BLS_FEDERAL_OtherSep_Level     20.402028
BLS_FEDERAL_Quits_Level        55.686488
BLS_FEDERAL_JobOpenings_Level   3.613370
BLS_FEDERAL_Layoffs_Rate       18.822912
BLS_FEDERAL_TotalSep_Rate      44.242917
SALARYLog                     103.348057
LOSSqrt                         1.952779
SEPCount_EFDATE_OCCLog          1.120721
SEPCount_EFDATE_LOCLog         22.379329
IndAvgSalaryLog               142.796845
AGELVL_B                        1.168674
AGELVL_C                        1.669532
AGELVL_E                        1.408634
AGELVL_F                        1.286868
AGELVL_G                        1.270860
AGELVL_H                        1.361018
AGELVL_I                        1.522204
LOC_01                          5.977821
LOC_02                          3.003188
LOC_04                          9.532952
LOC_05                          3.272509
LOC_06                         35.875014
LOC_08                         11.623086
LOC_09                          2.276701
LOC_10                          1.398734
LOC_11                         51.919874
LOC_12                         14.132643
LOC_13                         13.963587
LOC_15                          6.341591
LOC_16                          2.666040
LOC_17                          8.092630
LOC_18                          3.107775
LOC_19                          2.194186
LOC_20                          2.927057
LOC_21                          4.230129
LOC_22                          4.045685
LOC_23                          2.101691
LOC_24                         35.056214
LOC_25                          5.172459
LOC_26                          4.400721
LOC_27                          3.238691
LOC_28                          3.122100
LOC_29                          6.337389
LOC_30                          3.236332
LOC_31                          2.118288
LOC_32                          2.983542
LOC_33                          1.493756
LOC_34                          4.508410
LOC_35                          7.331036
LOC_36                          9.440269
LOC_37                          7.599446
LOC_38                          1.942321
LOC_39                          8.672915
LOC_40                          6.230374
LOC_41                          4.674291
LOC_42                          8.981797
LOC_44                          1.495404
LOC_45                          3.621864
LOC_46                          2.901286
LOC_47                          3.995110
LOC_48                         28.978212
LOC_49                          4.202065
LOC_50                          1.410012
LOC_51                         40.241437
LOC_53                         11.362128
LOC_54                          2.715366
LOC_55                          3.017254
TOA_10                         22.327307
TOA_15                         12.074569
TOA_20                          2.099988
TOA_30                          7.349533
TOA_32                          1.071774
TOA_35                          1.551110
TOA_38                         10.823001
TOA_40                          2.347558
TOA_42                          1.209745
TOA_44                          1.039621
TOA_45                          1.000003
PPGROUP_11                      1.882529


Removed BEFORE this step: IndAvgSalaryLog 

Call:
glm(formula = fla, family = "binomial", data = OPMAnalysisDataNoFamBinary)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-3.0395  -0.7019  -0.1263   0.7840   3.3497  

Coefficients:
                                Estimate Std. Error z value Pr(>|z|)    
(Intercept)                   -2.644e+01  4.770e+00  -5.545 2.95e-08 ***
GSEGRD                        -4.496e-01  3.992e-02 -11.262  < 2e-16 ***
IndAvgSalary                  -1.934e-06  3.952e-06  -0.489 0.624588    
SalaryOverUnderIndAvg         -2.292e-05  4.847e-06  -4.729 2.26e-06 ***
LowerLimitAge                 -5.021e-02  2.763e-03 -18.173  < 2e-16 ***
BLS_FEDERAL_OtherSep_Rate      1.430e+01  9.432e-01  15.157  < 2e-16 ***
BLS_FEDERAL_Quits_Rate        -2.875e+01  1.884e+00 -15.263  < 2e-16 ***
BLS_FEDERAL_OtherSep_Level    -3.027e-01  4.071e-02  -7.434 1.05e-13 ***
BLS_FEDERAL_Quits_Level        1.245e+00  7.710e-02  16.154  < 2e-16 ***
BLS_FEDERAL_JobOpenings_Level  6.282e-02  3.420e-03  18.368  < 2e-16 ***
BLS_FEDERAL_Layoffs_Rate       2.624e+00  4.567e-01   5.746 9.16e-09 ***
BLS_FEDERAL_TotalSep_Rate     -2.283e+00  4.730e-01  -4.826 1.39e-06 ***
SALARYLog                      2.781e+00  4.702e-01   5.915 3.33e-09 ***
LOSSqrt                       -7.773e-01  2.378e-02 -32.691  < 2e-16 ***
SEPCount_EFDATE_OCCLog        -4.269e-02  1.654e-02  -2.580 0.009867 ** 
SEPCount_EFDATE_LOCLog        -3.384e-01  1.206e-01  -2.806 0.005020 ** 
AGELVL_B1                     -7.060e-01  2.293e-01  -3.079 0.002077 ** 
AGELVL_C1                     -1.436e-01  8.760e-02  -1.639 0.101226    
AGELVL_E1                      2.734e-01  6.908e-02   3.957 7.59e-05 ***
AGELVL_F1                      4.953e-01  7.204e-02   6.877 6.13e-12 ***
AGELVL_G1                      7.354e-01  7.592e-02   9.687  < 2e-16 ***
AGELVL_H1                      8.824e-01  7.820e-02  11.284  < 2e-16 ***
AGELVL_I1                      7.463e-01  8.273e-02   9.020  < 2e-16 ***
LOC_011                       -1.231e-01  4.052e-01  -0.304 0.761344    
LOC_021                        3.745e-02  4.390e-01   0.085 0.932027    
LOC_041                        6.581e-01  4.085e-01   1.611 0.107178    
LOC_051                       -4.736e-02  4.248e-01  -0.111 0.911235    
LOC_061                        9.701e-01  4.771e-01   2.033 0.042035 *  
LOC_081                        3.861e-01  4.119e-01   0.937 0.348683    
LOC_091                       -5.468e-01  4.839e-01  -1.130 0.258549    
LOC_101                       -5.692e-01  6.976e-01  -0.816 0.414516    
LOC_111                        4.400e-01  4.626e-01   0.951 0.341499    
LOC_121                        3.965e-01  4.237e-01   0.936 0.349410    
LOC_131                        4.694e-01  4.145e-01   1.132 0.257439    
LOC_151                       -1.503e-01  3.971e-01  -0.379 0.705010    
LOC_161                       -6.249e-02  4.480e-01  -0.139 0.889065    
LOC_171                       -9.810e-02  4.077e-01  -0.241 0.809840    
LOC_181                       -2.354e-01  4.357e-01  -0.540 0.589030    
LOC_191                       -3.600e-02  4.806e-01  -0.075 0.940290    
LOC_201                        5.954e-01  4.383e-01   1.358 0.174363    
LOC_211                        3.674e-02  4.146e-01   0.089 0.929387    
LOC_221                       -2.066e-01  4.122e-01  -0.501 0.616202    
LOC_231                       -1.102e-01  4.970e-01  -0.222 0.824502    
LOC_241                        3.256e-01  4.395e-01   0.741 0.458834    
LOC_251                        1.379e-01  4.089e-01   0.337 0.736005    
LOC_261                       -5.016e-02  4.149e-01  -0.121 0.903763    
LOC_271                        2.662e-01  4.322e-01   0.616 0.537984    
LOC_281                       -1.075e-01  4.328e-01  -0.248 0.803895    
LOC_291                       -8.541e-02  4.248e-01  -0.201 0.840659    
LOC_301                        5.917e-01  4.322e-01   1.369 0.170951    
LOC_311                       -4.903e-02  4.859e-01  -0.101 0.919613    
LOC_321                        1.125e-01  4.350e-01   0.258 0.796022    
LOC_331                       -2.489e-01  6.473e-01  -0.385 0.700605    
LOC_341                       -2.017e-01  4.095e-01  -0.493 0.622246    
LOC_351                        5.241e-01  3.984e-01   1.316 0.188281    
LOC_361                        2.596e-01  4.252e-01   0.610 0.541536    
LOC_371                        3.067e-01  4.114e-01   0.746 0.455957    
LOC_381                        1.984e-01  5.107e-01   0.389 0.697620    
LOC_391                       -4.654e-02  4.083e-01  -0.114 0.909247    
LOC_401                        3.003e-01  4.034e-01   0.744 0.456696    
LOC_411                        2.491e-01  4.208e-01   0.592 0.553866    
LOC_421                       -2.432e-02  4.222e-01  -0.058 0.954061    
LOC_441                       -3.420e-01  6.440e-01  -0.531 0.595415    
LOC_451                        1.121e-02  4.221e-01   0.027 0.978806    
LOC_461                        5.917e-01  4.363e-01   1.356 0.175092    
LOC_471                       -1.381e-01  4.261e-01  -0.324 0.745863    
LOC_481                        8.442e-01  4.584e-01   1.842 0.065522 .  
LOC_491                        8.617e-03  4.239e-01   0.020 0.983780    
LOC_501                       -6.481e-01  6.984e-01  -0.928 0.353458    
LOC_511                        2.791e-01  4.526e-01   0.617 0.537366    
LOC_531                        6.714e-01  4.194e-01   1.601 0.109410    
LOC_541                       -2.588e-01  4.456e-01  -0.581 0.561371    
LOC_551                       -3.886e-01  4.369e-01  -0.889 0.373785    
TOA_101                       -1.102e+00  2.033e-01  -5.421 5.93e-08 ***
TOA_151                       -1.027e+00  2.083e-01  -4.928 8.32e-07 ***
TOA_201                        7.005e-02  2.747e-01   0.255 0.798677    
TOA_301                       -7.392e-01  2.187e-01  -3.379 0.000727 ***
TOA_321                       -4.529e-01  8.135e-01  -0.557 0.577697    
TOA_351                       -1.611e+00  3.531e-01  -4.563 5.05e-06 ***
TOA_381                       -7.100e-01  2.088e-01  -3.400 0.000673 ***
TOA_401                       -1.065e+00  2.652e-01  -4.016 5.92e-05 ***
TOA_421                        1.088e-01  4.845e-01   0.225 0.822336    
TOA_441                        9.136e-01  1.100e+00   0.831 0.406155    
TOA_451                        8.078e+00  1.343e+02   0.060 0.952049    
PPGROUP_111                    3.841e-01  1.524e-01   2.521 0.011706 *  
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 20683  on 14919  degrees of freedom
Residual deviance: 13940  on 14835  degrees of freedom
AIC: 14110

Number of Fisher Scoring iterations: 10

                                    VIF
GSEGRD                        10.387833
IndAvgSalary                  31.307265
SalaryOverUnderIndAvg          5.136218
LowerLimitAge                  2.394977
BLS_FEDERAL_OtherSep_Rate     13.403790
BLS_FEDERAL_Quits_Rate        38.696489
BLS_FEDERAL_OtherSep_Level    20.390819
BLS_FEDERAL_Quits_Level       55.669333
BLS_FEDERAL_JobOpenings_Level  3.612427
BLS_FEDERAL_Layoffs_Rate      18.792428
BLS_FEDERAL_TotalSep_Rate     44.170526
SALARYLog                     47.439530
LOSSqrt                        1.933243
SEPCount_EFDATE_OCCLog         1.119350
SEPCount_EFDATE_LOCLog        22.326922
AGELVL_B                       1.170956
AGELVL_C                       1.666558
AGELVL_E                       1.408466
AGELVL_F                       1.286235
AGELVL_G                       1.271210
AGELVL_H                       1.361442
AGELVL_I                       1.523344
LOC_01                         5.952585
LOC_02                         2.962028
LOC_04                         9.449940
LOC_05                         3.261701
LOC_06                        35.284461
LOC_08                        11.610517
LOC_09                         2.242291
LOC_10                         1.404120
LOC_11                        51.198890
LOC_12                        14.028929
LOC_13                        13.999906
LOC_15                         6.267715
LOC_16                         2.666982
LOC_17                         8.026589
LOC_18                         3.097305
LOC_19                         2.186868
LOC_20                         2.931389
LOC_21                         4.233893
LOC_22                         4.012930
LOC_23                         2.079279
LOC_24                        34.534464
LOC_25                         5.089867
LOC_26                         4.343076
LOC_27                         3.214594
LOC_28                         3.118743
LOC_29                         6.347257
LOC_30                         3.257016
LOC_31                         2.115996
LOC_32                         2.959568
LOC_33                         1.479543
LOC_34                         4.446378
LOC_35                         7.270815
LOC_36                         9.326787
LOC_37                         7.551486
LOC_38                         1.943499
LOC_39                         8.604161
LOC_40                         6.211647
LOC_41                         4.641935
LOC_42                         8.929351
LOC_44                         1.492800
LOC_45                         3.606664
LOC_46                         2.914152
LOC_47                         3.976128
LOC_48                        28.823356
LOC_49                         4.162500
LOC_50                         1.403730
LOC_51                        40.191955
LOC_53                        11.266994
LOC_54                         2.711282
LOC_55                         2.996195
TOA_10                        22.434899
TOA_15                        12.162260
TOA_20                         2.107040
TOA_30                         7.345564
TOA_32                         1.071489
TOA_35                         1.556871
TOA_38                        10.854026
TOA_40                         2.345046
TOA_42                         1.210866
TOA_44                         1.039612
TOA_45                         1.000003
PPGROUP_11                     1.811161


Removed BEFORE this step: BLS_FEDERAL_Quits_Level 

Call:
glm(formula = fla, family = "binomial", data = OPMAnalysisDataNoFamBinary)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-2.9018  -0.7194  -0.1324   0.8061   3.3221  

Coefficients:
                                Estimate Std. Error z value Pr(>|z|)    
(Intercept)                   -2.234e+01  4.713e+00  -4.740 2.14e-06 ***
GSEGRD                        -4.359e-01  3.959e-02 -11.011  < 2e-16 ***
IndAvgSalary                  -3.036e-06  3.907e-06  -0.777 0.437199    
SalaryOverUnderIndAvg         -2.260e-05  4.810e-06  -4.699 2.61e-06 ***
LowerLimitAge                 -4.989e-02  2.733e-03 -18.259  < 2e-16 ***
BLS_FEDERAL_OtherSep_Rate      8.371e+00  8.470e-01   9.884  < 2e-16 ***
BLS_FEDERAL_Quits_Rate         4.560e-02  5.967e-01   0.076 0.939079    
BLS_FEDERAL_OtherSep_Level    -3.959e-01  3.951e-02 -10.020  < 2e-16 ***
BLS_FEDERAL_JobOpenings_Level  2.186e-02  2.260e-03   9.673  < 2e-16 ***
BLS_FEDERAL_Layoffs_Rate      -2.568e+00  3.273e-01  -7.847 4.24e-15 ***
BLS_FEDERAL_TotalSep_Rate      3.397e+00  3.216e-01  10.564  < 2e-16 ***
SALARYLog                      2.747e+00  4.657e-01   5.898 3.67e-09 ***
LOSSqrt                       -7.801e-01  2.357e-02 -33.097  < 2e-16 ***
SEPCount_EFDATE_OCCLog        -3.452e-02  1.634e-02  -2.112 0.034675 *  
SEPCount_EFDATE_LOCLog        -4.259e-01  1.189e-01  -3.581 0.000342 ***
AGELVL_B1                     -6.924e-01  2.268e-01  -3.053 0.002265 ** 
AGELVL_C1                     -1.313e-01  8.679e-02  -1.513 0.130359    
AGELVL_E1                      2.763e-01  6.826e-02   4.047 5.18e-05 ***
AGELVL_F1                      4.965e-01  7.115e-02   6.978 2.99e-12 ***
AGELVL_G1                      7.324e-01  7.498e-02   9.768  < 2e-16 ***
AGELVL_H1                      8.746e-01  7.728e-02  11.318  < 2e-16 ***
AGELVL_I1                      7.430e-01  8.201e-02   9.060  < 2e-16 ***
LOC_011                       -5.601e-02  3.999e-01  -0.140 0.888592    
LOC_021                        5.742e-02  4.337e-01   0.132 0.894672    
LOC_041                        7.592e-01  4.037e-01   1.881 0.060021 .  
LOC_051                       -5.827e-02  4.196e-01  -0.139 0.889549    
LOC_061                        1.181e+00  4.712e-01   2.507 0.012167 *  
LOC_081                        4.987e-01  4.071e-01   1.225 0.220606    
LOC_091                       -5.831e-01  4.824e-01  -1.209 0.226831    
LOC_101                       -5.733e-01  6.939e-01  -0.826 0.408690    
LOC_111                        6.625e-01  4.565e-01   1.451 0.146646    
LOC_121                        5.224e-01  4.184e-01   1.249 0.211797    
LOC_131                        5.565e-01  4.093e-01   1.360 0.173940    
LOC_151                       -1.345e-01  3.927e-01  -0.342 0.732000    
LOC_161                       -9.205e-02  4.434e-01  -0.208 0.835545    
LOC_171                        3.706e-02  4.025e-01   0.092 0.926642    
LOC_181                       -2.620e-01  4.319e-01  -0.607 0.544081    
LOC_191                       -8.922e-02  4.737e-01  -0.188 0.850606    
LOC_201                        5.594e-01  4.332e-01   1.291 0.196549    
LOC_211                        3.146e-02  4.106e-01   0.077 0.938932    
LOC_221                       -1.711e-01  4.074e-01  -0.420 0.674576    
LOC_231                       -2.614e-01  4.908e-01  -0.533 0.594349    
LOC_241                        4.839e-01  4.337e-01   1.116 0.264584    
LOC_251                        1.387e-01  4.046e-01   0.343 0.731702    
LOC_261                       -6.759e-02  4.111e-01  -0.164 0.869401    
LOC_271                        2.541e-01  4.276e-01   0.594 0.552435    
LOC_281                       -8.418e-02  4.287e-01  -0.196 0.844337    
LOC_291                        1.340e-02  4.193e-01   0.032 0.974500    
LOC_301                        6.609e-01  4.286e-01   1.542 0.123099    
LOC_311                       -5.048e-02  4.834e-01  -0.104 0.916821    
LOC_321                        7.871e-02  4.283e-01   0.184 0.854173    
LOC_331                       -3.510e-01  6.392e-01  -0.549 0.582960    
LOC_341                       -2.261e-01  4.054e-01  -0.558 0.577028    
LOC_351                        5.660e-01  3.939e-01   1.437 0.150762    
LOC_361                        3.725e-01  4.195e-01   0.888 0.374574    
LOC_371                        3.360e-01  4.063e-01   0.827 0.408266    
LOC_381                        1.128e-01  5.042e-01   0.224 0.822940    
LOC_391                        1.617e-02  4.036e-01   0.040 0.968049    
LOC_401                        3.390e-01  3.991e-01   0.850 0.395598    
LOC_411                        3.394e-01  4.159e-01   0.816 0.414509    
LOC_421                        1.121e-01  4.167e-01   0.269 0.787960    
LOC_441                       -5.137e-01  6.313e-01  -0.814 0.415793    
LOC_451                        2.444e-02  4.183e-01   0.058 0.953414    
LOC_461                        5.432e-01  4.329e-01   1.255 0.209539    
LOC_471                       -5.328e-02  4.208e-01  -0.127 0.899227    
LOC_481                        9.828e-01  4.522e-01   2.173 0.029768 *  
LOC_491                        2.927e-02  4.189e-01   0.070 0.944286    
LOC_501                       -9.052e-01  6.916e-01  -1.309 0.190582    
LOC_511                        4.560e-01  4.466e-01   1.021 0.307159    
LOC_531                        7.725e-01  4.143e-01   1.865 0.062218 .  
LOC_541                       -2.808e-01  4.407e-01  -0.637 0.524045    
LOC_551                       -3.654e-01  4.315e-01  -0.847 0.397159    
TOA_101                       -1.071e+00  1.998e-01  -5.360 8.31e-08 ***
TOA_151                       -1.006e+00  2.047e-01  -4.911 9.04e-07 ***
TOA_201                        1.431e-01  2.725e-01   0.525 0.599441    
TOA_301                       -7.371e-01  2.152e-01  -3.425 0.000614 ***
TOA_321                       -6.133e-01  7.978e-01  -0.769 0.442008    
TOA_351                       -1.635e+00  3.487e-01  -4.690 2.73e-06 ***
TOA_381                       -6.999e-01  2.053e-01  -3.409 0.000651 ***
TOA_401                       -1.081e+00  2.610e-01  -4.141 3.47e-05 ***
TOA_421                        1.781e-01  4.812e-01   0.370 0.711334    
TOA_441                        1.043e+00  1.093e+00   0.955 0.339649    
TOA_451                        8.391e+00  1.372e+02   0.061 0.951241    
PPGROUP_111                    2.885e-01  1.503e-01   1.919 0.054947 .  
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 20683  on 14919  degrees of freedom
Residual deviance: 14209  on 14836  degrees of freedom
AIC: 14377

Number of Fisher Scoring iterations: 10

                                    VIF
GSEGRD                        10.424568
IndAvgSalary                  31.216174
SalaryOverUnderIndAvg          5.125810
LowerLimitAge                  2.387892
BLS_FEDERAL_OtherSep_Rate     10.875194
BLS_FEDERAL_Quits_Rate         4.193359
BLS_FEDERAL_OtherSep_Level    19.417270
BLS_FEDERAL_JobOpenings_Level  1.583558
BLS_FEDERAL_Layoffs_Rate       9.650140
BLS_FEDERAL_TotalSep_Rate     20.424399
SALARYLog                     47.466426
LOSSqrt                        1.927259
SEPCount_EFDATE_OCCLog         1.117936
SEPCount_EFDATE_LOCLog        22.153474
AGELVL_B                       1.171917
AGELVL_C                       1.662913
AGELVL_E                       1.408851
AGELVL_F                       1.288061
AGELVL_G                       1.274118
AGELVL_H                       1.364700
AGELVL_I                       1.523490
LOC_01                         6.047920
LOC_02                         2.980308
LOC_04                         9.395239
LOC_05                         3.281445
LOC_06                        35.067245
LOC_08                        11.530687
LOC_09                         2.205900
LOC_10                         1.401010
LOC_11                        50.768507
LOC_12                        13.990995
LOC_13                        13.927554
LOC_15                         6.250716
LOC_16                         2.659772
LOC_17                         8.087933
LOC_18                         3.066479
LOC_19                         2.206823
LOC_20                         2.938253
LOC_21                         4.188049
LOC_22                         4.026488
LOC_23                         2.091270
LOC_24                        34.578086
LOC_25                         5.046513
LOC_26                         4.281943
LOC_27                         3.203895
LOC_28                         3.099333
LOC_29                         6.320596
LOC_30                         3.228871
LOC_31                         2.091229
LOC_32                         3.019173
LOC_33                         1.482226
LOC_34                         4.407396
LOC_35                         7.197027
LOC_36                         9.405915
LOC_37                         7.613092
LOC_38                         1.950080
LOC_39                         8.532976
LOC_40                         6.158756
LOC_41                         4.640127
LOC_42                         9.008812
LOC_44                         1.506730
LOC_45                         3.562722
LOC_46                         2.885825
LOC_47                         3.986924
LOC_48                        28.862024
LOC_49                         4.170763
LOC_50                         1.402052
LOC_51                        40.072217
LOC_53                        11.336759
LOC_54                         2.712683
LOC_55                         3.013910
TOA_10                        22.095151
TOA_15                        11.934189
TOA_20                         2.059844
TOA_30                         7.207750
TOA_32                         1.071480
TOA_35                         1.548892
TOA_38                        10.705362
TOA_40                         2.331226
TOA_42                         1.203877
TOA_44                         1.038708
TOA_45                         1.000003
PPGROUP_11                     1.801222


Removed BEFORE this step: LOC_11 

Call:
glm(formula = fla, family = "binomial", data = OPMAnalysisDataNoFamBinary)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-2.9028  -0.7186  -0.1333   0.8045   3.3205  

Coefficients:
                                Estimate Std. Error z value Pr(>|z|)    
(Intercept)                   -2.261e+01  4.712e+00  -4.798 1.60e-06 ***
GSEGRD                        -4.352e-01  3.958e-02 -10.996  < 2e-16 ***
IndAvgSalary                  -3.102e-06  3.907e-06  -0.794 0.427148    
SalaryOverUnderIndAvg         -2.260e-05  4.810e-06  -4.698 2.62e-06 ***
LowerLimitAge                 -5.000e-02  2.732e-03 -18.303  < 2e-16 ***
BLS_FEDERAL_OtherSep_Rate      8.438e+00  8.457e-01   9.978  < 2e-16 ***
BLS_FEDERAL_Quits_Rate         4.259e-02  5.967e-01   0.071 0.943098    
BLS_FEDERAL_OtherSep_Level    -4.018e-01  3.931e-02 -10.221  < 2e-16 ***
BLS_FEDERAL_JobOpenings_Level  2.270e-02  2.185e-03  10.392  < 2e-16 ***
BLS_FEDERAL_Layoffs_Rate      -2.560e+00  3.272e-01  -7.825 5.06e-15 ***
BLS_FEDERAL_TotalSep_Rate      3.381e+00  3.213e-01  10.523  < 2e-16 ***
SALARYLog                      2.760e+00  4.656e-01   5.928 3.07e-09 ***
LOSSqrt                       -7.805e-01  2.358e-02 -33.108  < 2e-16 ***
SEPCount_EFDATE_OCCLog        -3.382e-02  1.633e-02  -2.071 0.038382 *  
SEPCount_EFDATE_LOCLog        -3.192e-01  9.371e-02  -3.406 0.000658 ***
AGELVL_B1                     -6.893e-01  2.268e-01  -3.039 0.002372 ** 
AGELVL_C1                     -1.320e-01  8.679e-02  -1.521 0.128343    
AGELVL_E1                      2.757e-01  6.826e-02   4.039 5.37e-05 ***
AGELVL_F1                      4.942e-01  7.112e-02   6.949 3.68e-12 ***
AGELVL_G1                      7.318e-01  7.497e-02   9.761  < 2e-16 ***
AGELVL_H1                      8.753e-01  7.727e-02  11.328  < 2e-16 ***
AGELVL_I1                      7.427e-01  8.201e-02   9.057  < 2e-16 ***
LOC_011                       -5.429e-01  2.186e-01  -2.484 0.012986 *  
LOC_021                       -3.766e-01  3.152e-01  -1.195 0.232105    
LOC_041                        2.316e-01  1.765e-01   1.312 0.189466    
LOC_051                       -4.678e-01  3.118e-01  -1.500 0.133599    
LOC_061                        5.149e-01  1.053e-01   4.891 1.00e-06 ***
LOC_081                       -4.657e-02  1.577e-01  -0.295 0.767768    
LOC_091                       -9.159e-01  4.259e-01  -2.150 0.031536 *  
LOC_101                       -8.526e-01  6.668e-01  -1.279 0.200988    
LOC_121                       -4.949e-02  1.412e-01  -0.350 0.726003    
LOC_131                        8.769e-04  1.456e-01   0.006 0.995194    
LOC_151                       -6.004e-01  2.275e-01  -2.639 0.008317 ** 
LOC_161                       -5.090e-01  3.385e-01  -1.504 0.132691    
LOC_171                       -4.812e-01  1.867e-01  -2.577 0.009977 ** 
LOC_181                       -7.062e-01  3.059e-01  -2.309 0.020943 *  
LOC_191                       -4.538e-01  4.032e-01  -1.126 0.260362    
LOC_201                        1.330e-01  3.194e-01   0.417 0.677038    
LOC_211                       -4.292e-01  2.617e-01  -1.640 0.100952    
LOC_221                       -6.101e-01  2.742e-01  -2.225 0.026071 *  
LOC_231                       -6.030e-01  4.321e-01  -1.396 0.162826    
LOC_241                       -1.295e-01  9.790e-02  -1.323 0.185760    
LOC_251                       -3.378e-01  2.376e-01  -1.422 0.155029    
LOC_261                       -5.353e-01  2.563e-01  -2.089 0.036734 *  
LOC_271                       -1.917e-01  2.988e-01  -0.641 0.521268    
LOC_281                       -5.197e-01  3.074e-01  -1.691 0.090904 .  
LOC_291                       -5.223e-01  1.996e-01  -2.616 0.008895 ** 
LOC_301                        2.088e-01  2.956e-01   0.706 0.479972    
LOC_311                       -4.305e-01  4.074e-01  -1.057 0.290665    
LOC_321                       -3.357e-01  3.206e-01  -1.047 0.295054    
LOC_331                       -6.502e-01  6.070e-01  -1.071 0.284063    
LOC_341                       -6.764e-01  2.621e-01  -2.581 0.009858 ** 
LOC_351                        7.981e-02  2.084e-01   0.383 0.701730    
LOC_361                       -1.877e-01  1.652e-01  -1.136 0.255979    
LOC_371                       -1.861e-01  1.899e-01  -0.980 0.327027    
LOC_381                       -2.361e-01  4.447e-01  -0.531 0.595422    
LOC_391                       -5.073e-01  1.821e-01  -2.786 0.005332 ** 
LOC_401                       -1.477e-01  2.174e-01  -0.679 0.496845    
LOC_411                       -1.571e-01  2.374e-01  -0.662 0.508030    
LOC_421                       -4.407e-01  1.697e-01  -2.597 0.009401 ** 
LOC_441                       -8.120e-01  5.983e-01  -1.357 0.174682    
LOC_451                       -4.226e-01  2.842e-01  -1.487 0.136989    
LOC_461                        1.526e-01  3.404e-01   0.448 0.654044    
LOC_471                       -5.370e-01  2.577e-01  -2.084 0.037174 *  
LOC_481                        3.449e-01  1.066e-01   3.237 0.001210 ** 
LOC_491                       -4.575e-01  2.521e-01  -1.815 0.069516 .  
LOC_501                       -1.188e+00  6.641e-01  -1.789 0.073561 .  
LOC_511                       -1.779e-01  9.329e-02  -1.907 0.056570 .  
LOC_531                        2.146e-01  1.553e-01   1.382 0.166843    
LOC_541                       -6.873e-01  3.415e-01  -2.013 0.044165 *  
LOC_551                       -7.985e-01  3.130e-01  -2.551 0.010744 *  
TOA_101                       -1.074e+00  1.998e-01  -5.374 7.68e-08 ***
TOA_151                       -1.007e+00  2.047e-01  -4.918 8.73e-07 ***
TOA_201                        1.419e-01  2.725e-01   0.521 0.602561    
TOA_301                       -7.353e-01  2.152e-01  -3.417 0.000633 ***
TOA_321                       -6.111e-01  7.974e-01  -0.766 0.443466    
TOA_351                       -1.632e+00  3.487e-01  -4.680 2.87e-06 ***
TOA_381                       -7.054e-01  2.053e-01  -3.436 0.000590 ***
TOA_401                       -1.081e+00  2.610e-01  -4.141 3.45e-05 ***
TOA_421                        1.774e-01  4.812e-01   0.369 0.712308    
TOA_441                        1.048e+00  1.093e+00   0.960 0.337239    
TOA_451                        8.400e+00  1.373e+02   0.061 0.951210    
PPGROUP_111                    2.932e-01  1.503e-01   1.951 0.051042 .  
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 20683  on 14919  degrees of freedom
Residual deviance: 14211  on 14837  degrees of freedom
AIC: 14377

Number of Fisher Scoring iterations: 10

                                    VIF
GSEGRD                        10.423477
IndAvgSalary                  31.212140
SalaryOverUnderIndAvg          5.124258
LowerLimitAge                  2.386331
BLS_FEDERAL_OtherSep_Rate     10.848400
BLS_FEDERAL_Quits_Rate         4.197164
BLS_FEDERAL_OtherSep_Level    19.227923
BLS_FEDERAL_JobOpenings_Level  1.480324
BLS_FEDERAL_Layoffs_Rate       9.649270
BLS_FEDERAL_TotalSep_Rate     20.404316
SALARYLog                     47.446166
LOSSqrt                        1.928047
SEPCount_EFDATE_OCCLog         1.116822
SEPCount_EFDATE_LOCLog        13.746535
AGELVL_B                       1.171734
AGELVL_C                       1.662693
AGELVL_E                       1.408708
AGELVL_F                       1.287400
AGELVL_G                       1.274043
AGELVL_H                       1.364665
AGELVL_I                       1.523507
LOC_01                         1.808712
LOC_02                         1.573336
LOC_04                         1.794369
LOC_05                         1.813621
LOC_06                         1.750672
LOC_08                         1.731481
LOC_09                         1.718502
LOC_10                         1.298869
LOC_12                         1.595397
LOC_13                         1.763800
LOC_15                         2.096169
LOC_16                         1.553767
LOC_17                         1.742101
LOC_18                         1.538529
LOC_19                         1.594760
LOC_20                         1.597676
LOC_21                         1.700998
LOC_22                         1.824250
LOC_23                         1.618931
LOC_24                         1.761633
LOC_25                         1.740938
LOC_26                         1.665517
LOC_27                         1.563316
LOC_28                         1.592336
LOC_29                         1.435470
LOC_30                         1.533660
LOC_31                         1.486334
LOC_32                         1.689900
LOC_33                         1.331779
LOC_34                         1.844282
LOC_35                         2.013207
LOC_36                         1.458929
LOC_37                         1.663214
LOC_38                         1.514536
LOC_39                         1.737623
LOC_40                         1.828184
LOC_41                         1.512959
LOC_42                         1.495241
LOC_44                         1.350909
LOC_45                         1.645623
LOC_46                         1.782745
LOC_47                         1.497655
LOC_48                         1.603522
LOC_49                         1.508097
LOC_50                         1.295152
LOC_51                         1.749001
LOC_53                         1.592749
LOC_54                         1.628118
LOC_55                         1.584264
TOA_10                        22.096299
TOA_15                        11.935872
TOA_20                         2.059526
TOA_30                         7.208034
TOA_32                         1.071549
TOA_35                         1.549060
TOA_38                        10.701220
TOA_40                         2.331845
TOA_42                         1.203958
TOA_44                         1.038700
TOA_45                         1.000003
PPGROUP_11                     1.799377


Removed BEFORE this step: SALARYLog 

Call:
glm(formula = fla, family = "binomial", data = OPMAnalysisDataNoFamBinary)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-2.9122  -0.7204  -0.1343   0.8061   3.2925  

Coefficients:
                                Estimate Std. Error z value Pr(>|z|)    
(Intercept)                    4.987e+00  7.835e-01   6.364 1.96e-10 ***
GSEGRD                        -2.955e-01  3.080e-02  -9.594  < 2e-16 ***
IndAvgSalary                   1.735e-05  1.898e-06   9.139  < 2e-16 ***
SalaryOverUnderIndAvg          2.150e-06  2.501e-06   0.860 0.389879    
LowerLimitAge                 -4.920e-02  2.725e-03 -18.051  < 2e-16 ***
BLS_FEDERAL_OtherSep_Rate      8.403e+00  8.448e-01   9.948  < 2e-16 ***
BLS_FEDERAL_Quits_Rate         1.083e-01  5.959e-01   0.182 0.855824    
BLS_FEDERAL_OtherSep_Level    -4.021e-01  3.928e-02 -10.236  < 2e-16 ***
BLS_FEDERAL_JobOpenings_Level  2.275e-02  2.183e-03  10.423  < 2e-16 ***
BLS_FEDERAL_Layoffs_Rate      -2.551e+00  3.267e-01  -7.808 5.79e-15 ***
BLS_FEDERAL_TotalSep_Rate      3.375e+00  3.209e-01  10.518  < 2e-16 ***
LOSSqrt                       -7.692e-01  2.344e-02 -32.816  < 2e-16 ***
SEPCount_EFDATE_OCCLog        -3.335e-02  1.631e-02  -2.044 0.040935 *  
SEPCount_EFDATE_LOCLog        -3.133e-01  9.411e-02  -3.329 0.000872 ***
AGELVL_B1                     -7.679e-01  2.251e-01  -3.411 0.000646 ***
AGELVL_C1                     -1.606e-01  8.665e-02  -1.854 0.063762 .  
AGELVL_E1                      2.733e-01  6.827e-02   4.004 6.24e-05 ***
AGELVL_F1                      4.870e-01  7.105e-02   6.855 7.14e-12 ***
AGELVL_G1                      7.267e-01  7.486e-02   9.708  < 2e-16 ***
AGELVL_H1                      8.658e-01  7.702e-02  11.240  < 2e-16 ***
AGELVL_I1                      7.354e-01  8.171e-02   9.000  < 2e-16 ***
LOC_011                       -5.557e-01  2.197e-01  -2.530 0.011416 *  
LOC_021                       -3.108e-01  3.146e-01  -0.988 0.323285    
LOC_041                        2.687e-01  1.765e-01   1.522 0.127996    
LOC_051                       -4.659e-01  3.134e-01  -1.487 0.137119    
LOC_061                        5.599e-01  1.046e-01   5.353 8.66e-08 ***
LOC_081                       -3.527e-02  1.582e-01  -0.223 0.823590    
LOC_091                       -8.671e-01  4.267e-01  -2.032 0.042120 *  
LOC_101                       -8.265e-01  6.705e-01  -1.233 0.217680    
LOC_121                       -5.599e-02  1.417e-01  -0.395 0.692810    
LOC_131                        1.182e-02  1.458e-01   0.081 0.935377    
LOC_151                       -5.282e-01  2.274e-01  -2.323 0.020202 *  
LOC_161                       -5.101e-01  3.394e-01  -1.503 0.132880    
LOC_171                       -4.700e-01  1.873e-01  -2.509 0.012105 *  
LOC_181                       -7.294e-01  3.070e-01  -2.375 0.017530 *  
LOC_191                       -4.637e-01  4.050e-01  -1.145 0.252203    
LOC_201                        1.018e-01  3.209e-01   0.317 0.751116    
LOC_211                       -4.880e-01  2.630e-01  -1.855 0.063583 .  
LOC_221                       -6.270e-01  2.759e-01  -2.273 0.023048 *  
LOC_231                       -5.873e-01  4.336e-01  -1.354 0.175613    
LOC_241                       -1.037e-01  9.743e-02  -1.064 0.287105    
LOC_251                       -2.889e-01  2.379e-01  -1.214 0.224606    
LOC_261                       -5.339e-01  2.576e-01  -2.073 0.038204 *  
LOC_271                       -1.689e-01  2.997e-01  -0.563 0.573110    
LOC_281                       -5.542e-01  3.093e-01  -1.792 0.073106 .  
LOC_291                       -5.440e-01  2.005e-01  -2.713 0.006663 ** 
LOC_301                        2.251e-01  2.959e-01   0.761 0.446735    
LOC_311                       -4.729e-01  4.088e-01  -1.157 0.247348    
LOC_321                       -3.299e-01  3.203e-01  -1.030 0.303020    
LOC_331                       -6.732e-01  6.067e-01  -1.110 0.267190    
LOC_341                       -6.022e-01  2.624e-01  -2.295 0.021717 *  
LOC_351                        8.758e-02  2.089e-01   0.419 0.675016    
LOC_361                       -1.700e-01  1.655e-01  -1.027 0.304265    
LOC_371                       -1.678e-01  1.903e-01  -0.882 0.377821    
LOC_381                       -2.658e-01  4.461e-01  -0.596 0.551256    
LOC_391                       -5.167e-01  1.826e-01  -2.830 0.004652 ** 
LOC_401                       -1.585e-01  2.183e-01  -0.726 0.467640    
LOC_411                       -1.645e-01  2.382e-01  -0.691 0.489675    
LOC_421                       -4.383e-01  1.703e-01  -2.573 0.010082 *  
LOC_441                       -7.260e-01  5.975e-01  -1.215 0.224288    
LOC_451                       -4.328e-01  2.856e-01  -1.516 0.129621    
LOC_461                        1.962e-01  3.412e-01   0.575 0.565204    
LOC_471                       -5.533e-01  2.581e-01  -2.144 0.032070 *  
LOC_481                        3.532e-01  1.064e-01   3.318 0.000907 ***
LOC_491                       -4.828e-01  2.539e-01  -1.901 0.057268 .  
LOC_501                       -1.227e+00  6.667e-01  -1.840 0.065731 .  
LOC_511                       -1.381e-01  9.239e-02  -1.494 0.135075    
LOC_531                        2.534e-01  1.554e-01   1.631 0.102945    
LOC_541                       -7.024e-01  3.430e-01  -2.048 0.040586 *  
LOC_551                       -8.292e-01  3.149e-01  -2.633 0.008458 ** 
TOA_101                       -1.032e+00  2.005e-01  -5.146 2.66e-07 ***
TOA_151                       -9.837e-01  2.057e-01  -4.783 1.73e-06 ***
TOA_201                        1.385e-01  2.737e-01   0.506 0.612667    
TOA_301                       -7.417e-01  2.158e-01  -3.437 0.000588 ***
TOA_321                       -6.169e-01  8.017e-01  -0.769 0.441625    
TOA_351                       -1.733e+00  3.489e-01  -4.968 6.76e-07 ***
TOA_381                       -6.428e-01  2.059e-01  -3.121 0.001800 ** 
TOA_401                       -1.089e+00  2.615e-01  -4.163 3.15e-05 ***
TOA_421                        1.516e-01  4.816e-01   0.315 0.752882    
TOA_441                        9.702e-01  1.092e+00   0.889 0.374127    
TOA_451                        8.342e+00  1.373e+02   0.061 0.951546    
PPGROUP_111                    3.996e-01  1.470e-01   2.718 0.006574 ** 
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 20683  on 14919  degrees of freedom
Residual deviance: 14247  on 14838  degrees of freedom
AIC: 14411

Number of Fisher Scoring iterations: 10

                                    VIF
GSEGRD                         6.354746
IndAvgSalary                   7.387766
SalaryOverUnderIndAvg          1.390114
LowerLimitAge                  2.385147
BLS_FEDERAL_OtherSep_Rate     10.845686
BLS_FEDERAL_Quits_Rate         4.193261
BLS_FEDERAL_OtherSep_Level    19.238975
BLS_FEDERAL_JobOpenings_Level  1.480661
BLS_FEDERAL_Layoffs_Rate       9.659385
BLS_FEDERAL_TotalSep_Rate     20.421006
LOSSqrt                        1.918764
SEPCount_EFDATE_OCCLog         1.115914
SEPCount_EFDATE_LOCLog        13.892472
AGELVL_B                       1.175315
AGELVL_C                       1.657862
AGELVL_E                       1.408735
AGELVL_F                       1.286049
AGELVL_G                       1.273136
AGELVL_H                       1.363132
AGELVL_I                       1.522601
LOC_01                         1.807478
LOC_02                         1.581945
LOC_04                         1.801385
LOC_05                         1.812283
LOC_06                         1.737309
LOC_08                         1.732556
LOC_09                         1.726621
LOC_10                         1.299121
LOC_12                         1.591573
LOC_13                         1.766830
LOC_15                         2.108753
LOC_16                         1.556842
LOC_17                         1.743444
LOC_18                         1.539210
LOC_19                         1.596619
LOC_20                         1.597037
LOC_21                         1.696238
LOC_22                         1.820706
LOC_23                         1.621595
LOC_24                         1.759641
LOC_25                         1.744307
LOC_26                         1.664777
LOC_27                         1.566618
LOC_28                         1.590589
LOC_29                         1.434045
LOC_30                         1.541124
LOC_31                         1.488876
LOC_32                         1.701623
LOC_33                         1.337304
LOC_34                         1.846826
LOC_35                         2.022638
LOC_36                         1.457910
LOC_37                         1.665035
LOC_38                         1.516009
LOC_39                         1.738870
LOC_40                         1.831806
LOC_41                         1.513169
LOC_42                         1.494536
LOC_44                         1.355905
LOC_45                         1.645585
LOC_46                         1.789284
LOC_47                         1.502473
LOC_48                         1.596570
LOC_49                         1.503876
LOC_50                         1.296301
LOC_51                         1.757573
LOC_53                         1.591402
LOC_54                         1.630070
LOC_55                         1.581435
TOA_10                        22.326646
TOA_15                        12.110061
TOA_20                         2.062102
TOA_30                         7.350634
TOA_32                         1.071336
TOA_35                         1.551092
TOA_38                        10.732947
TOA_40                         2.346874
TOA_42                         1.205710
TOA_44                         1.038804
TOA_45                         1.000003
PPGROUP_11                     1.750610


Removed BEFORE this step: TOA_10 

Call:
glm(formula = fla, family = "binomial", data = OPMAnalysisDataNoFamBinary)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-2.9215  -0.7214  -0.1332   0.8072   3.3048  

Coefficients:
                                Estimate Std. Error z value Pr(>|z|)    
(Intercept)                    4.041e+00  7.608e-01   5.312 1.09e-07 ***
GSEGRD                        -2.963e-01  3.065e-02  -9.667  < 2e-16 ***
IndAvgSalary                   1.719e-05  1.893e-06   9.082  < 2e-16 ***
SalaryOverUnderIndAvg          2.365e-06  2.502e-06   0.945 0.344447    
LowerLimitAge                 -4.878e-02  2.721e-03 -17.931  < 2e-16 ***
BLS_FEDERAL_OtherSep_Rate      8.451e+00  8.439e-01  10.014  < 2e-16 ***
BLS_FEDERAL_Quits_Rate         1.676e-01  5.943e-01   0.282 0.777969    
BLS_FEDERAL_OtherSep_Level    -4.039e-01  3.923e-02 -10.296  < 2e-16 ***
BLS_FEDERAL_JobOpenings_Level  2.261e-02  2.180e-03  10.368  < 2e-16 ***
BLS_FEDERAL_Layoffs_Rate      -2.514e+00  3.261e-01  -7.708 1.28e-14 ***
BLS_FEDERAL_TotalSep_Rate      3.350e+00  3.204e-01  10.457  < 2e-16 ***
LOSSqrt                       -7.863e-01  2.325e-02 -33.822  < 2e-16 ***
SEPCount_EFDATE_OCCLog        -3.478e-02  1.630e-02  -2.133 0.032932 *  
SEPCount_EFDATE_LOCLog        -3.075e-01  9.419e-02  -3.265 0.001095 ** 
AGELVL_B1                     -7.477e-01  2.245e-01  -3.330 0.000867 ***
AGELVL_C1                     -1.597e-01  8.638e-02  -1.848 0.064542 .  
AGELVL_E1                      2.741e-01  6.807e-02   4.027 5.65e-05 ***
AGELVL_F1                      4.863e-01  7.094e-02   6.855 7.14e-12 ***
AGELVL_G1                      7.226e-01  7.478e-02   9.663  < 2e-16 ***
AGELVL_H1                      8.614e-01  7.707e-02  11.176  < 2e-16 ***
AGELVL_I1                      7.289e-01  8.179e-02   8.911  < 2e-16 ***
LOC_011                       -5.653e-01  2.202e-01  -2.567 0.010247 *  
LOC_021                       -3.169e-01  3.146e-01  -1.007 0.313817    
LOC_041                        2.542e-01  1.765e-01   1.440 0.149857    
LOC_051                       -4.773e-01  3.138e-01  -1.521 0.128211    
LOC_061                        5.547e-01  1.044e-01   5.313 1.08e-07 ***
LOC_081                       -4.846e-02  1.582e-01  -0.306 0.759344    
LOC_091                       -8.718e-01  4.276e-01  -2.039 0.041488 *  
LOC_101                       -8.238e-01  6.711e-01  -1.228 0.219609    
LOC_121                       -5.109e-02  1.416e-01  -0.361 0.718248    
LOC_131                        9.091e-03  1.457e-01   0.062 0.950251    
LOC_151                       -5.436e-01  2.275e-01  -2.389 0.016876 *  
LOC_161                       -5.247e-01  3.399e-01  -1.544 0.122659    
LOC_171                       -4.739e-01  1.869e-01  -2.535 0.011230 *  
LOC_181                       -7.130e-01  3.068e-01  -2.324 0.020147 *  
LOC_191                       -4.297e-01  4.040e-01  -1.063 0.287556    
LOC_201                        9.455e-02  3.213e-01   0.294 0.768535    
LOC_211                       -5.051e-01  2.633e-01  -1.918 0.055063 .  
LOC_221                       -6.215e-01  2.763e-01  -2.249 0.024513 *  
LOC_231                       -6.024e-01  4.343e-01  -1.387 0.165415    
LOC_241                       -1.242e-01  9.734e-02  -1.276 0.202111    
LOC_251                       -2.691e-01  2.372e-01  -1.135 0.256559    
LOC_261                       -5.050e-01  2.566e-01  -1.968 0.049118 *  
LOC_271                       -1.794e-01  2.998e-01  -0.598 0.549698    
LOC_281                       -5.411e-01  3.080e-01  -1.757 0.078962 .  
LOC_291                       -5.541e-01  2.001e-01  -2.770 0.005612 ** 
LOC_301                        2.159e-01  2.957e-01   0.730 0.465262    
LOC_311                       -4.906e-01  4.089e-01  -1.200 0.230181    
LOC_321                       -3.230e-01  3.195e-01  -1.011 0.311986    
LOC_331                       -6.771e-01  6.082e-01  -1.113 0.265596    
LOC_341                       -6.184e-01  2.626e-01  -2.355 0.018502 *  
LOC_351                        7.983e-02  2.089e-01   0.382 0.702298    
LOC_361                       -1.493e-01  1.648e-01  -0.906 0.364981    
LOC_371                       -1.811e-01  1.903e-01  -0.952 0.341153    
LOC_381                       -2.808e-01  4.466e-01  -0.629 0.529406    
LOC_391                       -5.041e-01  1.824e-01  -2.764 0.005708 ** 
LOC_401                       -1.792e-01  2.185e-01  -0.820 0.412064    
LOC_411                       -1.674e-01  2.379e-01  -0.704 0.481595    
LOC_421                       -4.464e-01  1.702e-01  -2.623 0.008709 ** 
LOC_441                       -6.328e-01  5.863e-01  -1.079 0.280484    
LOC_451                       -4.241e-01  2.848e-01  -1.489 0.136417    
LOC_461                        1.782e-01  3.416e-01   0.522 0.601878    
LOC_471                       -5.485e-01  2.576e-01  -2.129 0.033224 *  
LOC_481                        3.360e-01  1.064e-01   3.159 0.001581 ** 
LOC_491                       -4.834e-01  2.534e-01  -1.907 0.056459 .  
LOC_501                       -1.231e+00  6.670e-01  -1.846 0.064920 .  
LOC_511                       -1.624e-01  9.230e-02  -1.759 0.078519 .  
LOC_531                        2.424e-01  1.553e-01   1.561 0.118452    
LOC_541                       -7.017e-01  3.438e-01  -2.041 0.041254 *  
LOC_551                       -8.288e-01  3.142e-01  -2.638 0.008331 ** 
TOA_151                       -2.428e-03  7.029e-02  -0.035 0.972449    
TOA_201                        1.125e+00  1.938e-01   5.802 6.54e-09 ***
TOA_301                        2.622e-01  8.984e-02   2.919 0.003515 ** 
TOA_321                        3.993e-01  7.766e-01   0.514 0.607112    
TOA_351                       -7.729e-01  2.923e-01  -2.645 0.008180 ** 
TOA_381                        3.477e-01  6.840e-02   5.083 3.72e-07 ***
TOA_401                       -1.069e-01  1.765e-01  -0.606 0.544827    
TOA_421                        1.134e+00  4.412e-01   2.569 0.010193 *  
TOA_441                        1.973e+00  1.074e+00   1.837 0.066271 .  
TOA_451                        9.288e+00  1.373e+02   0.068 0.946067    
PPGROUP_111                    3.754e-01  1.468e-01   2.558 0.010542 *  
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 20683  on 14919  degrees of freedom
Residual deviance: 14278  on 14839  degrees of freedom
AIC: 14440

Number of Fisher Scoring iterations: 10

                                    VIF
GSEGRD                         6.325307
IndAvgSalary                   7.356269
SalaryOverUnderIndAvg          1.389707
LowerLimitAge                  2.380904
BLS_FEDERAL_OtherSep_Rate     10.826082
BLS_FEDERAL_Quits_Rate         4.183464
BLS_FEDERAL_OtherSep_Level    19.198400
BLS_FEDERAL_JobOpenings_Level  1.481026
BLS_FEDERAL_Layoffs_Rate       9.606578
BLS_FEDERAL_TotalSep_Rate     20.328387
LOSSqrt                        1.885675
SEPCount_EFDATE_OCCLog         1.115541
SEPCount_EFDATE_LOCLog        13.944539
AGELVL_B                       1.172784
AGELVL_C                       1.658781
AGELVL_E                       1.407933
AGELVL_F                       1.285416
AGELVL_G                       1.272661
AGELVL_H                       1.362794
AGELVL_I                       1.520713
LOC_01                         1.804571
LOC_02                         1.585242
LOC_04                         1.803346
LOC_05                         1.811844
LOC_06                         1.739463
LOC_08                         1.732912
LOC_09                         1.724140
LOC_10                         1.299397
LOC_12                         1.593953
LOC_13                         1.770420
LOC_15                         2.113054
LOC_16                         1.555436
LOC_17                         1.751175
LOC_18                         1.541328
LOC_19                         1.603914
LOC_20                         1.596656
LOC_21                         1.696337
LOC_22                         1.819867
LOC_23                         1.620731
LOC_24                         1.756401
LOC_25                         1.753849
LOC_26                         1.675057
LOC_27                         1.568157
LOC_28                         1.600069
LOC_29                         1.437365
LOC_30                         1.540415
LOC_31                         1.490014
LOC_32                         1.707584
LOC_33                         1.336069
LOC_34                         1.847719
LOC_35                         2.028201
LOC_36                         1.465620
LOC_37                         1.666503
LOC_38                         1.516084
LOC_39                         1.744169
LOC_40                         1.831185
LOC_41                         1.516018
LOC_42                         1.496738
LOC_44                         1.372866
LOC_45                         1.653945
LOC_46                         1.788853
LOC_47                         1.507251
LOC_48                         1.595040
LOC_49                         1.508619
LOC_50                         1.296781
LOC_51                         1.752218
LOC_53                         1.593111
LOC_54                         1.627931
LOC_55                         1.587431
TOA_15                         1.412211
TOA_20                         1.029348
TOA_30                         1.266604
TOA_32                         1.006039
TOA_35                         1.086640
TOA_38                         1.179225
TOA_40                         1.064876
TOA_42                         1.011671
TOA_44                         1.005267
TOA_45                         1.000001
PPGROUP_11                     1.742598


Removed BEFORE this step: BLS_FEDERAL_TotalSep_Rate 

Call:
glm(formula = fla, family = "binomial", data = OPMAnalysisDataNoFamBinary)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-2.9353  -0.7302  -0.1356   0.8100   3.2309  

Coefficients:
                                Estimate Std. Error z value Pr(>|z|)    
(Intercept)                    3.674e+00  7.593e-01   4.839 1.30e-06 ***
GSEGRD                        -2.933e-01  3.058e-02  -9.591  < 2e-16 ***
IndAvgSalary                   1.695e-05  1.888e-06   8.979  < 2e-16 ***
SalaryOverUnderIndAvg          2.796e-06  2.492e-06   1.122 0.261873    
LowerLimitAge                 -4.835e-02  2.707e-03 -17.860  < 2e-16 ***
BLS_FEDERAL_OtherSep_Rate      9.281e+00  8.376e-01  11.080  < 2e-16 ***
BLS_FEDERAL_Quits_Rate         3.504e+00  5.004e-01   7.002 2.52e-12 ***
BLS_FEDERAL_OtherSep_Level    -2.781e-01  3.715e-02  -7.484 7.19e-14 ***
BLS_FEDERAL_JobOpenings_Level  1.762e-02  2.113e-03   8.338  < 2e-16 ***
BLS_FEDERAL_Layoffs_Rate       5.677e-01  1.403e-01   4.046 5.21e-05 ***
LOSSqrt                       -7.880e-01  2.314e-02 -34.047  < 2e-16 ***
SEPCount_EFDATE_OCCLog        -3.521e-02  1.623e-02  -2.170 0.029986 *  
SEPCount_EFDATE_LOCLog        -2.621e-01  9.394e-02  -2.791 0.005261 ** 
AGELVL_B1                     -7.745e-01  2.237e-01  -3.461 0.000537 ***
AGELVL_C1                     -1.638e-01  8.606e-02  -1.904 0.056965 .  
AGELVL_E1                      2.644e-01  6.777e-02   3.902 9.56e-05 ***
AGELVL_F1                      4.790e-01  7.067e-02   6.779 1.21e-11 ***
AGELVL_G1                      7.120e-01  7.434e-02   9.578  < 2e-16 ***
AGELVL_H1                      8.553e-01  7.668e-02  11.154  < 2e-16 ***
AGELVL_I1                      7.217e-01  8.141e-02   8.865  < 2e-16 ***
LOC_011                       -4.892e-01  2.192e-01  -2.232 0.025638 *  
LOC_021                       -2.776e-01  3.149e-01  -0.882 0.377925    
LOC_041                        2.839e-01  1.761e-01   1.612 0.106953    
LOC_051                       -4.012e-01  3.127e-01  -1.283 0.199506    
LOC_061                        5.450e-01  1.040e-01   5.239 1.61e-07 ***
LOC_081                       -2.602e-03  1.577e-01  -0.016 0.986836    
LOC_091                       -7.253e-01  4.261e-01  -1.702 0.088726 .  
LOC_101                       -6.300e-01  6.657e-01  -0.946 0.343986    
LOC_121                       -2.420e-02  1.414e-01  -0.171 0.864136    
LOC_131                        3.072e-02  1.451e-01   0.212 0.832282    
LOC_151                       -4.692e-01  2.275e-01  -2.063 0.039147 *  
LOC_161                       -4.091e-01  3.387e-01  -1.208 0.227138    
LOC_171                       -4.413e-01  1.862e-01  -2.370 0.017800 *  
LOC_181                       -6.293e-01  3.075e-01  -2.046 0.040713 *  
LOC_191                       -3.213e-01  4.049e-01  -0.794 0.427486    
LOC_201                        1.921e-01  3.208e-01   0.599 0.549229    
LOC_211                       -4.060e-01  2.617e-01  -1.551 0.120836    
LOC_221                       -5.259e-01  2.758e-01  -1.907 0.056546 .  
LOC_231                       -5.253e-01  4.323e-01  -1.215 0.224324    
LOC_241                       -1.295e-01  9.687e-02  -1.337 0.181206    
LOC_251                       -2.022e-01  2.357e-01  -0.858 0.391070    
LOC_261                       -4.227e-01  2.556e-01  -1.654 0.098182 .  
LOC_271                       -1.055e-01  2.964e-01  -0.356 0.722043    
LOC_281                       -4.927e-01  3.068e-01  -1.606 0.108340    
LOC_291                       -5.332e-01  1.989e-01  -2.681 0.007331 ** 
LOC_301                        2.638e-01  2.955e-01   0.893 0.371852    
LOC_311                       -4.621e-01  4.082e-01  -1.132 0.257680    
LOC_321                       -2.290e-01  3.172e-01  -0.722 0.470470    
LOC_331                       -4.697e-01  6.027e-01  -0.779 0.435789    
LOC_341                       -5.750e-01  2.621e-01  -2.194 0.028237 *  
LOC_351                        1.570e-01  2.079e-01   0.755 0.450005    
LOC_361                       -1.381e-01  1.638e-01  -0.843 0.399268    
LOC_371                       -1.477e-01  1.901e-01  -0.777 0.437210    
LOC_381                       -1.724e-01  4.481e-01  -0.385 0.700499    
LOC_391                       -4.344e-01  1.815e-01  -2.393 0.016691 *  
LOC_401                       -1.196e-01  2.180e-01  -0.548 0.583434    
LOC_411                       -1.232e-01  2.376e-01  -0.518 0.604112    
LOC_421                       -4.013e-01  1.694e-01  -2.369 0.017822 *  
LOC_441                       -3.872e-01  5.858e-01  -0.661 0.508579    
LOC_451                       -3.294e-01  2.843e-01  -1.159 0.246639    
LOC_461                        2.401e-01  3.400e-01   0.706 0.480184    
LOC_471                       -4.740e-01  2.563e-01  -1.849 0.064421 .  
LOC_481                        3.338e-01  1.059e-01   3.151 0.001627 ** 
LOC_491                       -4.321e-01  2.517e-01  -1.716 0.086092 .  
LOC_501                       -1.140e+00  6.713e-01  -1.698 0.089427 .  
LOC_511                       -1.533e-01  9.194e-02  -1.668 0.095375 .  
LOC_531                        2.774e-01  1.551e-01   1.789 0.073685 .  
LOC_541                       -5.950e-01  3.405e-01  -1.748 0.080527 .  
LOC_551                       -7.583e-01  3.127e-01  -2.425 0.015318 *  
TOA_151                       -1.144e-02  7.005e-02  -0.163 0.870254    
TOA_201                        1.091e+00  1.938e-01   5.631 1.79e-08 ***
TOA_301                        2.580e-01  8.931e-02   2.889 0.003869 ** 
TOA_321                        3.043e-01  7.653e-01   0.398 0.690928    
TOA_351                       -7.314e-01  2.925e-01  -2.501 0.012395 *  
TOA_381                        3.532e-01  6.815e-02   5.182 2.20e-07 ***
TOA_401                       -1.094e-01  1.758e-01  -0.622 0.533864    
TOA_421                        1.141e+00  4.406e-01   2.590 0.009598 ** 
TOA_441                        1.923e+00  1.075e+00   1.789 0.073605 .  
TOA_451                        9.298e+00  1.390e+02   0.067 0.946684    
PPGROUP_111                    3.646e-01  1.463e-01   2.491 0.012722 *  
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 20683  on 14919  degrees of freedom
Residual deviance: 14388  on 14840  degrees of freedom
AIC: 14548

Number of Fisher Scoring iterations: 10

                                    VIF
GSEGRD                         6.341945
IndAvgSalary                   7.370123
SalaryOverUnderIndAvg          1.389753
LowerLimitAge                  2.376196
BLS_FEDERAL_OtherSep_Rate     10.825917
BLS_FEDERAL_Quits_Rate         3.044004
BLS_FEDERAL_OtherSep_Level    17.475471
BLS_FEDERAL_JobOpenings_Level  1.407951
BLS_FEDERAL_Layoffs_Rate       1.799169
LOSSqrt                        1.879479
SEPCount_EFDATE_OCCLog         1.115124
SEPCount_EFDATE_LOCLog        13.998472
AGELVL_B                       1.174410
AGELVL_C                       1.657234
AGELVL_E                       1.408689
AGELVL_F                       1.284917
AGELVL_G                       1.273502
AGELVL_H                       1.363176
AGELVL_I                       1.520633
LOC_01                         1.811694
LOC_02                         1.585316
LOC_04                         1.800230
LOC_05                         1.819679
LOC_06                         1.738628
LOC_08                         1.736083
LOC_09                         1.726487
LOC_10                         1.303526
LOC_12                         1.591695
LOC_13                         1.779967
LOC_15                         2.118655
LOC_16                         1.562957
LOC_17                         1.756577
LOC_18                         1.534246
LOC_19                         1.596140
LOC_20                         1.595951
LOC_21                         1.703234
LOC_22                         1.817510
LOC_23                         1.628642
LOC_24                         1.762368
LOC_25                         1.760211
LOC_26                         1.678602
LOC_27                         1.583312
LOC_28                         1.604592
LOC_29                         1.443580
LOC_30                         1.539174
LOC_31                         1.491019
LOC_32                         1.718984
LOC_33                         1.340572
LOC_34                         1.852480
LOC_35                         2.032909
LOC_36                         1.473198
LOC_37                         1.665442
LOC_38                         1.508745
LOC_39                         1.748613
LOC_40                         1.834955
LOC_41                         1.517115
LOC_42                         1.500269
LOC_44                         1.370811
LOC_45                         1.653431
LOC_46                         1.798056
LOC_47                         1.514145
LOC_48                         1.593242
LOC_49                         1.518887
LOC_50                         1.291666
LOC_51                         1.752969
LOC_53                         1.594601
LOC_54                         1.645998
LOC_55                         1.592307
TOA_15                         1.409587
TOA_20                         1.028883
TOA_30                         1.267760
TOA_32                         1.006042
TOA_35                         1.086763
TOA_38                         1.179202
TOA_40                         1.064819
TOA_42                         1.011394
TOA_44                         1.005168
TOA_45                         1.000001
PPGROUP_11                     1.741409


Removed BEFORE this step: BLS_FEDERAL_OtherSep_Level 

Call:
glm(formula = fla, family = "binomial", data = OPMAnalysisDataNoFamBinary)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-2.9532  -0.7383  -0.1365   0.8198   3.1762  

Coefficients:
                                Estimate Std. Error z value Pr(>|z|)    
(Intercept)                    5.287e+00  7.319e-01   7.224 5.06e-13 ***
GSEGRD                        -2.922e-01  3.051e-02  -9.579  < 2e-16 ***
IndAvgSalary                   1.699e-05  1.884e-06   9.019  < 2e-16 ***
SalaryOverUnderIndAvg          3.106e-06  2.489e-06   1.248 0.211960    
LowerLimitAge                 -4.802e-02  2.703e-03 -17.763  < 2e-16 ***
BLS_FEDERAL_OtherSep_Rate      3.587e+00  3.444e-01  10.416  < 2e-16 ***
BLS_FEDERAL_Quits_Rate         6.748e-01  3.255e-01   2.073 0.038190 *  
BLS_FEDERAL_JobOpenings_Level  1.429e-02  2.067e-03   6.911 4.82e-12 ***
BLS_FEDERAL_Layoffs_Rate       8.280e-02  1.245e-01   0.665 0.506114    
LOSSqrt                       -7.882e-01  2.310e-02 -34.119  < 2e-16 ***
SEPCount_EFDATE_OCCLog        -3.684e-02  1.620e-02  -2.274 0.022949 *  
SEPCount_EFDATE_LOCLog        -3.727e-01  9.321e-02  -3.999 6.36e-05 ***
AGELVL_B1                     -7.646e-01  2.235e-01  -3.421 0.000624 ***
AGELVL_C1                     -1.656e-01  8.580e-02  -1.930 0.053636 .  
AGELVL_E1                      2.686e-01  6.757e-02   3.975 7.04e-05 ***
AGELVL_F1                      4.737e-01  7.044e-02   6.724 1.77e-11 ***
AGELVL_G1                      7.031e-01  7.420e-02   9.475  < 2e-16 ***
AGELVL_H1                      8.441e-01  7.650e-02  11.034  < 2e-16 ***
AGELVL_I1                      7.181e-01  8.133e-02   8.829  < 2e-16 ***
LOC_011                       -6.540e-01  2.187e-01  -2.991 0.002784 ** 
LOC_021                       -4.875e-01  3.135e-01  -1.555 0.119925    
LOC_041                        1.660e-01  1.753e-01   0.947 0.343661    
LOC_051                       -6.508e-01  3.112e-01  -2.091 0.036505 *  
LOC_061                        5.681e-01  1.038e-01   5.473 4.42e-08 ***
LOC_081                       -1.113e-01  1.572e-01  -0.708 0.479025    
LOC_091                       -1.049e+00  4.271e-01  -2.457 0.014018 *  
LOC_101                       -9.956e-01  6.615e-01  -1.505 0.132320    
LOC_121                       -9.499e-02  1.412e-01  -0.673 0.500990    
LOC_131                       -6.792e-02  1.445e-01  -0.470 0.638345    
LOC_151                       -6.324e-01  2.262e-01  -2.796 0.005168 ** 
LOC_161                       -6.305e-01  3.373e-01  -1.869 0.061607 .  
LOC_171                       -5.595e-01  1.854e-01  -3.017 0.002550 ** 
LOC_181                       -8.125e-01  3.072e-01  -2.644 0.008182 ** 
LOC_191                       -6.385e-01  4.020e-01  -1.588 0.112244    
LOC_201                       -3.906e-02  3.195e-01  -0.122 0.902696    
LOC_211                       -5.910e-01  2.609e-01  -2.265 0.023522 *  
LOC_221                       -7.343e-01  2.749e-01  -2.671 0.007554 ** 
LOC_231                       -8.506e-01  4.285e-01  -1.985 0.047143 *  
LOC_241                       -1.614e-01  9.661e-02  -1.671 0.094767 .  
LOC_251                       -3.625e-01  2.351e-01  -1.542 0.123038    
LOC_261                       -6.001e-01  2.544e-01  -2.359 0.018345 *  
LOC_271                       -3.339e-01  2.949e-01  -1.132 0.257591    
LOC_281                       -7.362e-01  3.052e-01  -2.412 0.015867 *  
LOC_291                       -6.371e-01  1.987e-01  -3.206 0.001344 ** 
LOC_301                        7.847e-02  2.940e-01   0.267 0.789516    
LOC_311                       -7.103e-01  4.067e-01  -1.747 0.080684 .  
LOC_321                       -4.641e-01  3.165e-01  -1.466 0.142534    
LOC_331                       -8.586e-01  5.959e-01  -1.441 0.149650    
LOC_341                       -7.725e-01  2.604e-01  -2.967 0.003006 ** 
LOC_351                       -3.262e-03  2.072e-01  -0.016 0.987437    
LOC_361                       -2.407e-01  1.634e-01  -1.473 0.140694    
LOC_371                       -2.692e-01  1.896e-01  -1.420 0.155614    
LOC_381                       -4.733e-01  4.474e-01  -1.058 0.290112    
LOC_391                       -5.567e-01  1.804e-01  -3.086 0.002030 ** 
LOC_401                       -2.741e-01  2.176e-01  -1.260 0.207833    
LOC_411                       -2.777e-01  2.369e-01  -1.172 0.241023    
LOC_421                       -4.883e-01  1.689e-01  -2.892 0.003829 ** 
LOC_441                       -7.413e-01  5.787e-01  -1.281 0.200190    
LOC_451                       -5.634e-01  2.829e-01  -1.992 0.046409 *  
LOC_461                       -3.944e-02  3.376e-01  -0.117 0.906993    
LOC_471                       -6.524e-01  2.550e-01  -2.558 0.010530 *  
LOC_481                        3.332e-01  1.056e-01   3.156 0.001599 ** 
LOC_491                       -5.904e-01  2.506e-01  -2.356 0.018468 *  
LOC_501                       -1.484e+00  6.731e-01  -2.206 0.027418 *  
LOC_511                       -1.649e-01  9.177e-02  -1.797 0.072338 .  
LOC_531                        1.878e-01  1.543e-01   1.217 0.223567    
LOC_541                       -8.106e-01  3.388e-01  -2.393 0.016726 *  
LOC_551                       -9.633e-01  3.115e-01  -3.092 0.001985 ** 
TOA_151                       -7.958e-03  6.983e-02  -0.114 0.909269    
TOA_201                        1.108e+00  1.936e-01   5.725 1.04e-08 ***
TOA_301                        2.554e-01  8.920e-02   2.863 0.004190 ** 
TOA_321                        3.097e-01  7.628e-01   0.406 0.684736    
TOA_351                       -7.413e-01  2.919e-01  -2.539 0.011107 *  
TOA_381                        3.605e-01  6.796e-02   5.304 1.13e-07 ***
TOA_401                       -9.923e-02  1.751e-01  -0.567 0.570842    
TOA_421                        1.154e+00  4.401e-01   2.621 0.008761 ** 
TOA_441                        1.955e+00  1.074e+00   1.820 0.068706 .  
TOA_451                        9.497e+00  1.389e+02   0.068 0.945507    
PPGROUP_111                    3.845e-01  1.460e-01   2.634 0.008441 ** 
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 20683  on 14919  degrees of freedom
Residual deviance: 14445  on 14841  degrees of freedom
AIC: 14603

Number of Fisher Scoring iterations: 10

                                    VIF
GSEGRD                         6.340938
IndAvgSalary                   7.356814
SalaryOverUnderIndAvg          1.391728
LowerLimitAge                  2.376786
BLS_FEDERAL_OtherSep_Rate      1.837442
BLS_FEDERAL_Quits_Rate         1.304298
BLS_FEDERAL_JobOpenings_Level  1.352153
BLS_FEDERAL_Layoffs_Rate       1.426971
LOSSqrt                        1.876778
SEPCount_EFDATE_OCCLog         1.114597
SEPCount_EFDATE_LOCLog        13.850861
AGELVL_B                       1.173733
AGELVL_C                       1.657051
AGELVL_E                       1.408001
AGELVL_F                       1.285024
AGELVL_G                       1.273653
AGELVL_H                       1.364834
AGELVL_I                       1.522927
LOC_01                         1.801446
LOC_02                         1.579994
LOC_04                         1.797805
LOC_05                         1.813437
LOC_06                         1.736764
LOC_08                         1.729533
LOC_09                         1.703796
LOC_10                         1.302073
LOC_12                         1.587882
LOC_13                         1.771815
LOC_15                         2.115854
LOC_16                         1.564891
LOC_17                         1.756185
LOC_18                         1.527946
LOC_19                         1.593072
LOC_20                         1.588630
LOC_21                         1.697363
LOC_22                         1.808539
LOC_23                         1.628683
LOC_24                         1.763482
LOC_25                         1.755846
LOC_26                         1.675467
LOC_27                         1.575193
LOC_28                         1.596194
LOC_29                         1.441176
LOC_30                         1.531651
LOC_31                         1.488323
LOC_32                         1.704091
LOC_33                         1.342222
LOC_34                         1.854575
LOC_35                         2.022229
LOC_36                         1.466953
LOC_37                         1.660334
LOC_38                         1.501093
LOC_39                         1.748446
LOC_40                         1.825332
LOC_41                         1.510419
LOC_42                         1.498880
LOC_44                         1.375688
LOC_45                         1.641557
LOC_46                         1.791183
LOC_47                         1.510186
LOC_48                         1.594814
LOC_49                         1.514824
LOC_50                         1.285312
LOC_51                         1.752417
LOC_53                         1.592484
LOC_54                         1.649127
LOC_55                         1.590688
TOA_15                         1.409930
TOA_20                         1.028661
TOA_30                         1.269052
TOA_32                         1.006115
TOA_35                         1.085718
TOA_38                         1.178584
TOA_40                         1.065235
TOA_42                         1.011066
TOA_44                         1.005141
TOA_45                         1.000001
PPGROUP_11                     1.732739


Removed BEFORE this step: SEPCount_EFDATE_LOCLog 

Call:
glm(formula = fla, family = "binomial", data = OPMAnalysisDataNoFamBinary)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-2.9404  -0.7343  -0.1383   0.8212   3.1759  

Coefficients:
                                Estimate Std. Error z value Pr(>|z|)    
(Intercept)                    2.745e+00  3.596e-01   7.635 2.26e-14 ***
GSEGRD                        -2.950e-01  3.049e-02  -9.675  < 2e-16 ***
IndAvgSalary                   1.689e-05  1.882e-06   8.974  < 2e-16 ***
SalaryOverUnderIndAvg          2.876e-06  2.485e-06   1.158 0.247034    
LowerLimitAge                 -4.756e-02  2.697e-03 -17.631  < 2e-16 ***
BLS_FEDERAL_OtherSep_Rate      3.346e+00  3.388e-01   9.875  < 2e-16 ***
BLS_FEDERAL_Quits_Rate         3.838e-01  3.169e-01   1.211 0.225844    
BLS_FEDERAL_JobOpenings_Level  1.720e-02  1.938e-03   8.877  < 2e-16 ***
BLS_FEDERAL_Layoffs_Rate       1.669e-02  1.233e-01   0.135 0.892372    
LOSSqrt                       -7.857e-01  2.306e-02 -34.069  < 2e-16 ***
SEPCount_EFDATE_OCCLog        -3.844e-02  1.618e-02  -2.376 0.017523 *  
AGELVL_B1                     -7.652e-01  2.235e-01  -3.424 0.000618 ***
AGELVL_C1                     -1.649e-01  8.578e-02  -1.923 0.054507 .  
AGELVL_E1                      2.708e-01  6.752e-02   4.011 6.06e-05 ***
AGELVL_F1                      4.758e-01  7.038e-02   6.759 1.38e-11 ***
AGELVL_G1                      7.042e-01  7.414e-02   9.498  < 2e-16 ***
AGELVL_H1                      8.438e-01  7.643e-02  11.040  < 2e-16 ***
AGELVL_I1                      7.158e-01  8.128e-02   8.807  < 2e-16 ***
LOC_011                       -1.307e-01  1.749e-01  -0.748 0.454749    
LOC_021                        2.189e-01  2.586e-01   0.846 0.397322    
LOC_041                        5.389e-01  1.483e-01   3.634 0.000279 ***
LOC_051                        1.374e-01  2.405e-01   0.571 0.567951    
LOC_061                        4.687e-01  1.007e-01   4.654 3.25e-06 ***
LOC_081                        2.092e-01  1.350e-01   1.549 0.121329    
LOC_091                        1.354e-02  3.349e-01   0.040 0.967746    
LOC_101                        2.475e-01  5.805e-01   0.426 0.669886    
LOC_121                        1.287e-01  1.294e-01   0.995 0.319910    
LOC_131                        2.155e-01  1.257e-01   1.714 0.086450 .  
LOC_151                       -2.843e-02  1.684e-01  -0.169 0.865939    
LOC_161                        1.264e-01  2.783e-01   0.454 0.649756    
LOC_171                       -1.401e-01  1.526e-01  -0.918 0.358403    
LOC_181                       -1.428e-01  2.571e-01  -0.555 0.578769    
LOC_191                        3.025e-01  3.269e-01   0.926 0.354679    
LOC_201                        6.877e-01  2.626e-01   2.619 0.008813 ** 
LOC_211                        2.026e-02  2.112e-01   0.096 0.923577    
LOC_221                       -4.942e-02  2.147e-01  -0.230 0.817933    
LOC_231                        1.774e-01  3.423e-01   0.518 0.604303    
LOC_241                       -6.758e-02  9.365e-02  -0.722 0.470546    
LOC_251                        2.047e-01  1.870e-01   1.095 0.273489    
LOC_261                       -1.108e-02  2.070e-01  -0.054 0.957324    
LOC_271                        3.257e-01  2.446e-01   1.332 0.182939    
LOC_281                       -4.111e-02  2.509e-01  -0.164 0.869845    
LOC_291                       -2.891e-01  1.781e-01  -1.624 0.104428    
LOC_301                        7.125e-01  2.483e-01   2.869 0.004112 ** 
LOC_311                        1.772e-01  3.401e-01   0.521 0.602346    
LOC_321                        3.018e-01  2.520e-01   1.198 0.230974    
LOC_331                        3.095e-01  5.209e-01   0.594 0.552413    
LOC_341                       -1.125e-01  2.008e-01  -0.560 0.575224    
LOC_351                        5.158e-01  1.615e-01   3.193 0.001407 ** 
LOC_361                        3.142e-02  1.484e-01   0.212 0.832277    
LOC_371                        1.303e-01  1.608e-01   0.810 0.417993    
LOC_381                        5.196e-01  3.731e-01   1.393 0.163639    
LOC_391                       -1.624e-01  1.508e-01  -1.077 0.281509    
LOC_401                        2.483e-01  1.738e-01   1.428 0.153223    
LOC_411                        2.023e-01  2.037e-01   0.993 0.320573    
LOC_421                       -1.907e-01  1.513e-01  -1.260 0.207503    
LOC_441                        4.373e-01  4.973e-01   0.879 0.379227    
LOC_451                        8.970e-02  2.304e-01   0.389 0.696994    
LOC_461                        8.073e-01  2.631e-01   3.069 0.002149 ** 
LOC_471                       -1.224e-01  2.173e-01  -0.563 0.573384    
LOC_481                        3.287e-01  1.055e-01   3.117 0.001830 ** 
LOC_491                       -7.061e-02  2.145e-01  -0.329 0.741976    
LOC_501                       -2.519e-01  5.972e-01  -0.422 0.673226    
LOC_511                       -1.460e-01  9.164e-02  -1.593 0.111186    
LOC_531                        4.629e-01  1.379e-01   3.357 0.000787 ***
LOC_541                       -1.260e-03  2.717e-01  -0.005 0.996301    
LOC_551                       -2.568e-01  2.565e-01  -1.001 0.316627    
TOA_151                       -9.509e-03  6.981e-02  -0.136 0.891646    
TOA_201                        1.114e+00  1.935e-01   5.756 8.60e-09 ***
TOA_301                        2.531e-01  8.918e-02   2.838 0.004544 ** 
TOA_321                        3.369e-01  7.593e-01   0.444 0.657309    
TOA_351                       -7.563e-01  2.916e-01  -2.594 0.009496 ** 
TOA_381                        3.576e-01  6.789e-02   5.268 1.38e-07 ***
TOA_401                       -1.062e-01  1.750e-01  -0.607 0.543981    
TOA_421                        1.145e+00  4.400e-01   2.603 0.009241 ** 
TOA_441                        1.927e+00  1.073e+00   1.795 0.072617 .  
TOA_451                        9.465e+00  1.390e+02   0.068 0.945714    
PPGROUP_111                    3.738e-01  1.457e-01   2.565 0.010330 *  
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 20683  on 14919  degrees of freedom
Residual deviance: 14461  on 14842  degrees of freedom
AIC: 14617

Number of Fisher Scoring iterations: 10

                                   VIF
GSEGRD                        6.336016
IndAvgSalary                  7.358771
SalaryOverUnderIndAvg         1.389651
LowerLimitAge                 2.369312
BLS_FEDERAL_OtherSep_Rate     1.782477
BLS_FEDERAL_Quits_Rate        1.239868
BLS_FEDERAL_JobOpenings_Level 1.188769
BLS_FEDERAL_Layoffs_Rate      1.402466
LOSSqrt                       1.872374
SEPCount_EFDATE_OCCLog        1.114156
AGELVL_B                      1.174018
AGELVL_C                      1.657118
AGELVL_E                      1.408279
AGELVL_F                      1.284967
AGELVL_G                      1.273409
AGELVL_H                      1.363993
AGELVL_I                      1.521333
LOC_01                        1.155460
LOC_02                        1.076294
LOC_04                        1.288291
LOC_05                        1.084919
LOC_06                        1.638853
LOC_08                        1.279200
LOC_09                        1.043171
LOC_10                        1.015127
LOC_12                        1.338156
LOC_13                        1.344786
LOC_15                        1.170698
LOC_16                        1.069523
LOC_17                        1.192634
LOC_18                        1.072685
LOC_19                        1.046542
LOC_20                        1.074673
LOC_21                        1.113178
LOC_22                        1.104323
LOC_23                        1.041955
LOC_24                        1.658068
LOC_25                        1.115819
LOC_26                        1.112352
LOC_27                        1.080944
LOC_28                        1.077166
LOC_29                        1.163211
LOC_30                        1.089820
LOC_31                        1.044267
LOC_32                        1.078497
LOC_33                        1.019745
LOC_34                        1.107036
LOC_35                        1.228142
LOC_36                        1.211008
LOC_37                        1.197596
LOC_38                        1.038371
LOC_39                        1.225100
LOC_40                        1.165699
LOC_41                        1.121614
LOC_42                        1.206579
LOC_44                        1.018926
LOC_45                        1.092520
LOC_46                        1.085858
LOC_47                        1.101845
LOC_48                        1.596293
LOC_49                        1.105336
LOC_50                        1.015236
LOC_51                        1.746972
LOC_53                        1.275738
LOC_54                        1.059452
LOC_55                        1.076786
TOA_15                        1.409602
TOA_20                        1.028455
TOA_30                        1.268027
TOA_32                        1.006043
TOA_35                        1.086146
TOA_38                        1.178145
TOA_40                        1.064906
TOA_42                        1.011026
TOA_44                        1.005096
TOA_45                        1.000001
PPGROUP_11                    1.734680

Following variables removed based on VIF values (in order of removal):
 [1] "BLS_FEDERAL_JobOpenings_Rate" "BLS_FEDERAL_TotalSep_Level"  
 [3] "BLS_FEDERAL_Layoffs_Level"    "AGELVL_D"                    
 [5] "IndAvgSalaryLog"              "BLS_FEDERAL_Quits_Level"     
 [7] "LOC_11"                       "SALARYLog"                   
 [9] "TOA_10"                       "BLS_FEDERAL_TotalSep_Rate"   
[11] "BLS_FEDERAL_OtherSep_Level"   "SEPCount_EFDATE_LOCLog"      


Null Deviances (in order):
 [1] "7226.05152246447" "7219.36370064357" "6767.88407188207" "6767.82808686825"
 [5] "6742.92826531162" "6473.82478843413" "6471.71501030948" "6436.43729999307"
 [9] "6405.40525464846" "6294.74008501623" "6238.61063516369" "6222.45281211915"

Min value at iteration =  12

Diff Degrees of Freedom (in order):
 [1] "88" "87" "86" "85" "84" "83" "82" "81" "80" "79" "78" "77"

Min value at iteration =  12

Log Likelihoods (in order):
 [1] "-6728.54665990542" "-6731.89057081586" "-6957.63038519662"
 [4] "-6957.65837770352" "-6970.10828848184" "-7104.66002692059"
 [7] "-7105.71491598291" "-7123.35377114112" "-7138.86979381342"
[10] "-7194.20237862954" "-7222.2671035558"  "-7230.34601507808"

Min value at iteration =  12

AIC values (in order):
 [1] "13635.0933198108" "13639.7811416317" "14089.2607703932" "14087.316755407" 
 [5] "14110.2165769637" "14377.3200538412" "14377.4298319658" "14410.7075422822"
 [9] "14439.7395876268" "14548.4047572591" "14602.5342071116" "14616.6920301562"

Min value at iteration =  1

BIC values (in order):
 [1] "14312.4240705753" "14309.5014345224" "14751.3706054102" "14741.8161325502"
 [5] "14757.1054962331" "15016.5985152368" "15009.0978354877" "15034.7650879304"
 [9] "15056.1866754012" "15157.2413871597" "15203.7603791385" "15210.3077443093"

Min value at iteration =  2

%%R
#vars.Repeat <- replace(vars.Repeat, vars.Repeat == "IndAvgSalary", "IndAvgSalaryLog")
vars.Repeat <- vars.Repeat[!vars.Repeat %in% c("IndAvgSalary", "IndAvgSalaryLog")]

%%R

##data.frame(summary(BinLogit)$coef[summary(BinLogit)$coef[,4] <= .05, 4]) #Review coefficients of p-value less than 0.05
#LogitCoeffs <- data.frame(summary(BinLogit.P.Repeat)$coef[-1,4]) #Ignore Intercept and only look at p-values
##LogitCoeffs[LogitCoeffs$`summary.BinLogit..coef..1..4.` == max(LogitCoeffs$`summary.BinLogit..coef..1..4.`),]
#maxP <- cbind(rownames(LogitCoeffs)[LogitCoeffs$`summary.BinLogit.P.Repeat..coef..1..4.` == max(LogitCoeffs$`summary.BinLogit.P.Repeat..coef..1..4.`)],
#      max(LogitCoeffs))
#maxP[1,1]

varsP.Repeat <- vars.Repeat

P.removed <- vector(mode="character", length=0)
ndev.vect <- vector(mode="character", length=0)
ndf.vect <- vector(mode="character", length=0)
pchisq.vect <- vector(mode="character", length=0)
logLik.vect <- vector(mode="character", length=0)
AIC.vect <- vector(mode="character", length=0)
BIC.vect <- vector(mode="character", length=0)

for(i in seq(1,45)){
    BinLogit.P.Repeat <- runLogit(remove, varsP.Repeat)
    print(summary(BinLogit.P.Repeat))
    
    LogitCoeffs <- data.frame(summary(BinLogit.P.Repeat)$coef[-1,4]) #Ignore Intercept and only look at p-values
    maxP <- cbind(rownames(LogitCoeffs)[LogitCoeffs$`summary.BinLogit.P.Repeat..coef..1..4.` == max(LogitCoeffs$`summary.BinLogit.P.Repeat..coef..1..4.`)],
          max(LogitCoeffs))
    
    vifs.BinLogitP.Repeat <- runVifs(BinLogit.P.Repeat)
    
    remove <- maxP[1,1]
    remove <- ifelse(grepl('[[:digit:]]$', remove), substr(remove, 1, nchar(remove)-1), remove)
    P.removed <- c(P.removed, remove)
    cat("\nRemoved AFTER this step:", remove, "\n\n\n")
    varsP.Repeat <- varsP.Repeat[!(varsP.Repeat %in% c(remove))]
    
    ##goodness of fit
    ndev.vect <- c(ndev.vect, with(BinLogit.P.Repeat, null.deviance - deviance))
    ndf.vect <- c(ndf.vect, with(BinLogit.P.Repeat, df.null - df.residual))
    pchisq.vect <- c(pchisq.vect, with(BinLogit.P.Repeat, pchisq(null.deviance - deviance, df.null - df.residual, lower.tail = FALSE)))
    logLik.vect <- c(logLik.vect, logLik(BinLogit.P.Repeat))
    AIC.vect <- c(AIC.vect, AIC(BinLogit.P.Repeat))
    BIC.vect <- c(BIC.vect, BIC(BinLogit.P.Repeat))
}

cat("\nFollowing variables removed based on p-values (in order of removal):\n")
print(P.removed)

cat("\n\nNull Deviances (in order):\n")
print(ndev.vect)
cat("\nMin value at iteration = ", which.min(ndev.vect))

cat("\n\nDiff Degrees of Freedom (in order):\n")
print(ndf.vect)
cat("\nMin value at iteration = ", which.min(ndf.vect))

#cat("\n\nP-ChiSquare (in order):\n")
#print(pchisq.vect)
#cat("\nMin value at iteration = ", which.min(pchisq.vect))

cat("\n\nLog Likelihoods (in order):\n")
print(logLik.vect)
cat("\nMin value at iteration = ", which.min(logLik.vect))

cat("\n\nAIC values (in order):\n")
print(AIC.vect)
cat("\nMin value at iteration = ", which.min(AIC.vect))

cat("\n\nBIC values (in order):\n")
print(BIC.vect)
cat("\nMin value at iteration = ", which.min(BIC.vect))

Call:
glm(formula = fla, family = "binomial", data = OPMAnalysisDataNoFamBinary)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-2.8257  -0.7389  -0.1369   0.8192   3.2349  

Coefficients:
                                Estimate Std. Error z value Pr(>|z|)    
(Intercept)                    2.164e+00  3.501e-01   6.180 6.40e-10 ***
GSEGRD                        -6.306e-02  1.543e-02  -4.086 4.38e-05 ***
SalaryOverUnderIndAvg          8.942e-07  2.407e-06   0.371 0.710304    
LowerLimitAge                 -4.492e-02  2.664e-03 -16.863  < 2e-16 ***
BLS_FEDERAL_OtherSep_Rate      3.241e+00  3.376e-01   9.600  < 2e-16 ***
BLS_FEDERAL_Quits_Rate         4.739e-01  3.158e-01   1.500 0.133506    
BLS_FEDERAL_JobOpenings_Level  1.707e-02  1.931e-03   8.841  < 2e-16 ***
BLS_FEDERAL_Layoffs_Rate       4.260e-02  1.230e-01   0.346 0.729124    
LOSSqrt                       -7.824e-01  2.293e-02 -34.117  < 2e-16 ***
SEPCount_EFDATE_OCCLog        -3.092e-02  1.612e-02  -1.919 0.055040 .  
AGELVL_B1                     -5.865e-01  2.200e-01  -2.666 0.007681 ** 
AGELVL_C1                     -1.685e-01  8.534e-02  -1.974 0.048329 *  
AGELVL_E1                      2.740e-01  6.732e-02   4.071 4.69e-05 ***
AGELVL_F1                      4.894e-01  7.015e-02   6.976 3.04e-12 ***
AGELVL_G1                      7.122e-01  7.401e-02   9.623  < 2e-16 ***
AGELVL_H1                      8.467e-01  7.623e-02  11.106  < 2e-16 ***
AGELVL_I1                      7.176e-01  8.111e-02   8.847  < 2e-16 ***
LOC_011                       -2.079e-01  1.746e-01  -1.191 0.233732    
LOC_021                        1.563e-01  2.560e-01   0.611 0.541422    
LOC_041                        5.289e-01  1.471e-01   3.595 0.000325 ***
LOC_051                        8.352e-02  2.384e-01   0.350 0.726039    
LOC_061                        4.123e-01  1.007e-01   4.096 4.21e-05 ***
LOC_081                        1.457e-01  1.347e-01   1.081 0.279502    
LOC_091                       -5.574e-02  3.326e-01  -0.168 0.866909    
LOC_101                        1.984e-01  5.833e-01   0.340 0.733717    
LOC_121                        3.653e-02  1.292e-01   0.283 0.777292    
LOC_131                        1.713e-01  1.257e-01   1.363 0.172839    
LOC_151                       -6.696e-02  1.679e-01  -0.399 0.690072    
LOC_161                        5.430e-02  2.754e-01   0.197 0.843696    
LOC_171                       -2.199e-01  1.517e-01  -1.450 0.147157    
LOC_181                       -2.260e-01  2.553e-01  -0.885 0.375953    
LOC_191                        2.173e-01  3.232e-01   0.672 0.501296    
LOC_201                        6.153e-01  2.607e-01   2.360 0.018276 *  
LOC_211                       -2.959e-02  2.090e-01  -0.142 0.887417    
LOC_221                       -1.098e-01  2.132e-01  -0.515 0.606432    
LOC_231                        7.089e-02  3.402e-01   0.208 0.834951    
LOC_241                       -1.077e-01  9.422e-02  -1.144 0.252826    
LOC_251                        1.454e-01  1.869e-01   0.778 0.436632    
LOC_261                       -7.013e-02  2.065e-01  -0.340 0.734203    
LOC_271                        2.898e-01  2.424e-01   1.196 0.231868    
LOC_281                       -1.234e-01  2.506e-01  -0.492 0.622505    
LOC_291                       -3.485e-01  1.763e-01  -1.978 0.047985 *  
LOC_301                        6.821e-01  2.453e-01   2.781 0.005425 ** 
LOC_311                        1.345e-01  3.381e-01   0.398 0.690894    
LOC_321                        2.332e-01  2.485e-01   0.939 0.347934    
LOC_331                        1.992e-01  5.206e-01   0.383 0.701933    
LOC_341                       -1.807e-01  2.003e-01  -0.902 0.367024    
LOC_351                        4.741e-01  1.597e-01   2.968 0.003000 ** 
LOC_361                       -2.760e-02  1.483e-01  -0.186 0.852409    
LOC_371                        6.922e-02  1.600e-01   0.433 0.665374    
LOC_381                        4.903e-01  3.698e-01   1.326 0.184807    
LOC_391                       -2.892e-01  1.504e-01  -1.922 0.054549 .  
LOC_401                        1.795e-01  1.726e-01   1.040 0.298214    
LOC_411                        1.169e-01  2.027e-01   0.577 0.563965    
LOC_421                       -2.936e-01  1.508e-01  -1.947 0.051510 .  
LOC_441                        3.399e-01  4.936e-01   0.689 0.490990    
LOC_451                        6.030e-02  2.297e-01   0.263 0.792925    
LOC_461                        8.158e-01  2.590e-01   3.149 0.001636 ** 
LOC_471                       -1.775e-01  2.153e-01  -0.824 0.409699    
LOC_481                        2.529e-01  1.052e-01   2.404 0.016200 *  
LOC_491                       -1.834e-01  2.140e-01  -0.857 0.391511    
LOC_501                       -4.076e-01  5.947e-01  -0.685 0.493070    
LOC_511                       -1.541e-01  9.176e-02  -1.679 0.093158 .  
LOC_531                        3.993e-01  1.373e-01   2.909 0.003627 ** 
LOC_541                       -6.049e-02  2.704e-01  -0.224 0.823000    
LOC_551                       -3.197e-01  2.548e-01  -1.254 0.209681    
TOA_151                        6.868e-02  6.912e-02   0.994 0.320443    
TOA_201                        1.128e+00  1.928e-01   5.850 4.91e-09 ***
TOA_301                        3.464e-01  8.928e-02   3.880 0.000104 ***
TOA_321                        4.203e-01  7.540e-01   0.557 0.577220    
TOA_351                       -5.348e-01  2.867e-01  -1.865 0.062116 .  
TOA_381                        3.287e-01  6.730e-02   4.884 1.04e-06 ***
TOA_401                       -1.054e-01  1.765e-01  -0.597 0.550292    
TOA_421                        1.141e+00  4.414e-01   2.585 0.009751 ** 
TOA_441                        2.165e+00  1.073e+00   2.018 0.043574 *  
TOA_451                        9.288e+00  1.390e+02   0.067 0.946731    
PPGROUP_111                   -4.096e-01  1.172e-01  -3.494 0.000477 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 20683  on 14919  degrees of freedom
Residual deviance: 14543  on 14843  degrees of freedom
AIC: 14697

Number of Fisher Scoring iterations: 10

                                   VIF
GSEGRD                        1.709771
SalaryOverUnderIndAvg         1.362063
LowerLimitAge                 2.331432
BLS_FEDERAL_OtherSep_Rate     1.780588
BLS_FEDERAL_Quits_Rate        1.239370
BLS_FEDERAL_JobOpenings_Level 1.189501
BLS_FEDERAL_Layoffs_Rate      1.400736
LOSSqrt                       1.863535
SEPCount_EFDATE_OCCLog        1.110247
AGELVL_B                      1.168836
AGELVL_C                      1.661736
AGELVL_E                      1.407665
AGELVL_F                      1.284149
AGELVL_G                      1.271136
AGELVL_H                      1.360827
AGELVL_I                      1.516939
LOC_01                        1.155470
LOC_02                        1.077857
LOC_04                        1.298072
LOC_05                        1.086904
LOC_06                        1.643628
LOC_08                        1.281373
LOC_09                        1.043573
LOC_10                        1.015121
LOC_12                        1.333413
LOC_13                        1.347233
LOC_15                        1.173350
LOC_16                        1.070565
LOC_17                        1.194540
LOC_18                        1.073599
LOC_19                        1.046991
LOC_20                        1.074966
LOC_21                        1.115080
LOC_22                        1.105564
LOC_23                        1.041794
LOC_24                        1.659934
LOC_25                        1.116609
LOC_26                        1.113438
LOC_27                        1.083260
LOC_28                        1.076607
LOC_29                        1.166602
LOC_30                        1.092269
LOC_31                        1.044699
LOC_32                        1.080552
LOC_33                        1.019291
LOC_34                        1.108298
LOC_35                        1.236012
LOC_36                        1.212675
LOC_37                        1.199467
LOC_38                        1.038974
LOC_39                        1.218402
LOC_40                        1.167278
LOC_41                        1.121708
LOC_42                        1.204208
LOC_44                        1.018967
LOC_45                        1.093473
LOC_46                        1.088792
LOC_47                        1.104474
LOC_48                        1.598139
LOC_49                        1.103074
LOC_50                        1.014634
LOC_51                        1.786297
LOC_53                        1.278948
LOC_54                        1.060266
LOC_55                        1.077488
TOA_15                        1.405927
TOA_20                        1.028658
TOA_30                        1.249601
TOA_32                        1.005946
TOA_35                        1.078285
TOA_38                        1.177160
TOA_40                        1.063196
TOA_42                        1.010740
TOA_44                        1.004523
TOA_45                        1.000001
PPGROUP_11                    1.158838

Removed AFTER this step: TOA_45 


Call:
glm(formula = fla, family = "binomial", data = OPMAnalysisDataNoFamBinary)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-2.8259  -0.7389  -0.1369   0.8190   3.2351  

Coefficients:
                                Estimate Std. Error z value Pr(>|z|)    
(Intercept)                    2.165e+00  3.501e-01   6.183 6.28e-10 ***
GSEGRD                        -6.310e-02  1.543e-02  -4.088 4.35e-05 ***
SalaryOverUnderIndAvg          9.006e-07  2.407e-06   0.374 0.708324    
LowerLimitAge                 -4.492e-02  2.664e-03 -16.860  < 2e-16 ***
BLS_FEDERAL_OtherSep_Rate      3.241e+00  3.376e-01   9.600  < 2e-16 ***
BLS_FEDERAL_Quits_Rate         4.753e-01  3.158e-01   1.505 0.132305    
BLS_FEDERAL_JobOpenings_Level  1.707e-02  1.931e-03   8.842  < 2e-16 ***
BLS_FEDERAL_Layoffs_Rate       4.242e-02  1.230e-01   0.345 0.730206    
LOSSqrt                       -7.826e-01  2.293e-02 -34.129  < 2e-16 ***
SEPCount_EFDATE_OCCLog        -3.101e-02  1.612e-02  -1.924 0.054362 .  
AGELVL_B1                     -5.865e-01  2.200e-01  -2.666 0.007678 ** 
AGELVL_C1                     -1.678e-01  8.533e-02  -1.966 0.049254 *  
AGELVL_E1                      2.741e-01  6.732e-02   4.071 4.68e-05 ***
AGELVL_F1                      4.894e-01  7.015e-02   6.976 3.03e-12 ***
AGELVL_G1                      7.122e-01  7.401e-02   9.623  < 2e-16 ***
AGELVL_H1                      8.467e-01  7.623e-02  11.107  < 2e-16 ***
AGELVL_I1                      7.176e-01  8.111e-02   8.847  < 2e-16 ***
LOC_011                       -2.086e-01  1.746e-01  -1.195 0.232083    
LOC_021                        1.555e-01  2.560e-01   0.608 0.543406    
LOC_041                        5.282e-01  1.471e-01   3.590 0.000331 ***
LOC_051                        8.277e-02  2.384e-01   0.347 0.728420    
LOC_061                        4.115e-01  1.007e-01   4.088 4.35e-05 ***
LOC_081                        1.449e-01  1.347e-01   1.076 0.281892    
LOC_091                       -5.656e-02  3.326e-01  -0.170 0.864989    
LOC_101                        1.977e-01  5.833e-01   0.339 0.734644    
LOC_121                        3.582e-02  1.291e-01   0.277 0.781519    
LOC_131                        1.706e-01  1.257e-01   1.358 0.174603    
LOC_151                       -6.775e-02  1.679e-01  -0.403 0.686601    
LOC_161                        5.356e-02  2.754e-01   0.194 0.845806    
LOC_171                       -2.207e-01  1.517e-01  -1.455 0.145722    
LOC_181                       -2.268e-01  2.553e-01  -0.888 0.374317    
LOC_191                        2.165e-01  3.232e-01   0.670 0.502848    
LOC_201                        6.146e-01  2.607e-01   2.357 0.018410 *  
LOC_211                       -3.030e-02  2.090e-01  -0.145 0.884738    
LOC_221                       -1.106e-01  2.132e-01  -0.519 0.603787    
LOC_231                        7.004e-02  3.403e-01   0.206 0.836914    
LOC_241                       -1.085e-01  9.421e-02  -1.152 0.249486    
LOC_251                        1.446e-01  1.869e-01   0.774 0.438976    
LOC_261                       -7.089e-02  2.065e-01  -0.343 0.731407    
LOC_271                        2.891e-01  2.424e-01   1.192 0.233090    
LOC_281                       -1.241e-01  2.506e-01  -0.495 0.620487    
LOC_291                       -3.493e-01  1.763e-01  -1.982 0.047514 *  
LOC_301                        6.813e-01  2.453e-01   2.778 0.005476 ** 
LOC_311                        1.336e-01  3.381e-01   0.395 0.692692    
LOC_321                        2.325e-01  2.485e-01   0.935 0.349541    
LOC_331                        1.985e-01  5.206e-01   0.381 0.703042    
LOC_341                       -1.816e-01  2.003e-01  -0.906 0.364685    
LOC_351                        4.734e-01  1.597e-01   2.963 0.003043 ** 
LOC_361                       -2.832e-02  1.483e-01  -0.191 0.848601    
LOC_371                        6.847e-02  1.600e-01   0.428 0.668753    
LOC_381                        4.895e-01  3.698e-01   1.324 0.185597    
LOC_391                       -2.900e-01  1.504e-01  -1.927 0.053932 .  
LOC_401                        1.787e-01  1.726e-01   1.036 0.300293    
LOC_411                        1.162e-01  2.027e-01   0.573 0.566526    
LOC_421                       -2.943e-01  1.508e-01  -1.952 0.050904 .  
LOC_441                        3.392e-01  4.936e-01   0.687 0.491901    
LOC_451                        5.953e-02  2.297e-01   0.259 0.795515    
LOC_461                        8.151e-01  2.590e-01   3.146 0.001653 ** 
LOC_471                       -1.783e-01  2.153e-01  -0.828 0.407746    
LOC_481                        2.522e-01  1.052e-01   2.398 0.016500 *  
LOC_491                       -1.842e-01  2.140e-01  -0.861 0.389502    
LOC_501                       -4.084e-01  5.947e-01  -0.687 0.492256    
LOC_511                       -1.548e-01  9.175e-02  -1.688 0.091467 .  
LOC_531                        3.985e-01  1.373e-01   2.903 0.003694 ** 
LOC_541                       -6.119e-02  2.704e-01  -0.226 0.820980    
LOC_551                       -3.205e-01  2.548e-01  -1.258 0.208523    
TOA_151                        6.828e-02  6.912e-02   0.988 0.323244    
TOA_201                        1.128e+00  1.928e-01   5.849 4.96e-09 ***
TOA_301                        3.461e-01  8.928e-02   3.877 0.000106 ***
TOA_321                        4.203e-01  7.540e-01   0.557 0.577220    
TOA_351                       -5.356e-01  2.867e-01  -1.868 0.061749 .  
TOA_381                        3.285e-01  6.731e-02   4.881 1.06e-06 ***
TOA_401                       -1.060e-01  1.765e-01  -0.601 0.548033    
TOA_421                        1.140e+00  4.414e-01   2.583 0.009781 ** 
TOA_441                        2.164e+00  1.073e+00   2.018 0.043636 *  
PPGROUP_111                   -4.097e-01  1.172e-01  -3.495 0.000475 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 20683  on 14919  degrees of freedom
Residual deviance: 14543  on 14844  degrees of freedom
AIC: 14695

Number of Fisher Scoring iterations: 5

                                   VIF
GSEGRD                        1.709771
SalaryOverUnderIndAvg         1.362036
LowerLimitAge                 2.331493
BLS_FEDERAL_OtherSep_Rate     1.780582
BLS_FEDERAL_Quits_Rate        1.239390
BLS_FEDERAL_JobOpenings_Level 1.189508
BLS_FEDERAL_Layoffs_Rate      1.400708
LOSSqrt                       1.863360
SEPCount_EFDATE_OCCLog        1.110277
AGELVL_B                      1.168845
AGELVL_C                      1.661983
AGELVL_E                      1.407668
AGELVL_F                      1.284155
AGELVL_G                      1.271141
AGELVL_H                      1.360831
AGELVL_I                      1.516931
LOC_01                        1.155373
LOC_02                        1.077814
LOC_04                        1.297941
LOC_05                        1.086855
LOC_06                        1.643160
LOC_08                        1.281207
LOC_09                        1.043546
LOC_10                        1.015115
LOC_12                        1.333230
LOC_13                        1.347045
LOC_15                        1.173237
LOC_16                        1.070530
LOC_17                        1.194398
LOC_18                        1.073553
LOC_19                        1.046963
LOC_20                        1.074933
LOC_21                        1.115026
LOC_22                        1.105494
LOC_23                        1.041765
LOC_24                        1.659471
LOC_25                        1.116529
LOC_26                        1.113368
LOC_27                        1.083215
LOC_28                        1.076568
LOC_29                        1.166510
LOC_30                        1.092223
LOC_31                        1.044672
LOC_32                        1.080506
LOC_33                        1.019281
LOC_34                        1.108202
LOC_35                        1.235907
LOC_36                        1.212547
LOC_37                        1.199355
LOC_38                        1.038948
LOC_39                        1.218269
LOC_40                        1.167172
LOC_41                        1.121634
LOC_42                        1.204074
LOC_44                        1.018958
LOC_45                        1.093417
LOC_46                        1.088753
LOC_47                        1.104412
LOC_48                        1.597803
LOC_49                        1.103006
LOC_50                        1.014627
LOC_51                        1.785718
LOC_53                        1.278745
LOC_54                        1.060235
LOC_55                        1.077441
TOA_15                        1.405743
TOA_20                        1.028644
TOA_30                        1.249555
TOA_32                        1.005946
TOA_35                        1.078261
TOA_38                        1.177114
TOA_40                        1.063136
TOA_42                        1.010736
TOA_44                        1.004521
PPGROUP_11                    1.158849

Removed AFTER this step: LOC_21 


Call:
glm(formula = fla, family = "binomial", data = OPMAnalysisDataNoFamBinary)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-2.8258  -0.7389  -0.1369   0.8188   3.2357  

Coefficients:
                                Estimate Std. Error z value Pr(>|z|)    
(Intercept)                    2.159e+00  3.482e-01   6.201 5.61e-10 ***
GSEGRD                        -6.284e-02  1.533e-02  -4.098 4.16e-05 ***
SalaryOverUnderIndAvg          9.164e-07  2.405e-06   0.381  0.70321    
LowerLimitAge                 -4.494e-02  2.661e-03 -16.889  < 2e-16 ***
BLS_FEDERAL_OtherSep_Rate      3.241e+00  3.376e-01   9.599  < 2e-16 ***
BLS_FEDERAL_Quits_Rate         4.759e-01  3.158e-01   1.507  0.13178    
BLS_FEDERAL_JobOpenings_Level  1.707e-02  1.931e-03   8.841  < 2e-16 ***
BLS_FEDERAL_Layoffs_Rate       4.248e-02  1.230e-01   0.345  0.72985    
LOSSqrt                       -7.826e-01  2.293e-02 -34.131  < 2e-16 ***
SEPCount_EFDATE_OCCLog        -3.111e-02  1.610e-02  -1.932  0.05334 .  
AGELVL_B1                     -5.861e-01  2.200e-01  -2.664  0.00772 ** 
AGELVL_C1                     -1.677e-01  8.533e-02  -1.966  0.04933 *  
AGELVL_E1                      2.740e-01  6.732e-02   4.070 4.70e-05 ***
AGELVL_F1                      4.894e-01  7.015e-02   6.976 3.03e-12 ***
AGELVL_G1                      7.121e-01  7.400e-02   9.622  < 2e-16 ***
AGELVL_H1                      8.466e-01  7.623e-02  11.106  < 2e-16 ***
AGELVL_I1                      7.177e-01  8.111e-02   8.849  < 2e-16 ***
LOC_011                       -2.055e-01  1.732e-01  -1.186  0.23554    
LOC_021                        1.588e-01  2.550e-01   0.623  0.53336    
LOC_041                        5.316e-01  1.452e-01   3.662  0.00025 ***
LOC_051                        8.602e-02  2.373e-01   0.363  0.71697    
LOC_061                        4.145e-01  9.845e-02   4.210 2.55e-05 ***
LOC_081                        1.481e-01  1.329e-01   1.114  0.26515    
LOC_091                       -5.332e-02  3.319e-01  -0.161  0.87236    
LOC_101                        2.011e-01  5.828e-01   0.345  0.73006    
LOC_121                        3.902e-02  1.272e-01   0.307  0.75911    
LOC_131                        1.738e-01  1.237e-01   1.405  0.16012    
LOC_151                       -6.454e-02  1.665e-01  -0.388  0.69819    
LOC_161                        5.689e-02  2.744e-01   0.207  0.83578    
LOC_171                       -2.177e-01  1.503e-01  -1.449  0.14744    
LOC_181                       -2.236e-01  2.544e-01  -0.879  0.37930    
LOC_191                        2.198e-01  3.224e-01   0.682  0.49538    
LOC_201                        6.179e-01  2.597e-01   2.380  0.01733 *  
LOC_221                       -1.074e-01  2.120e-01  -0.507  0.61248    
LOC_231                        7.316e-02  3.396e-01   0.215  0.82941    
LOC_241                       -1.056e-01  9.214e-02  -1.147  0.25156    
LOC_251                        1.476e-01  1.857e-01   0.794  0.42698    
LOC_261                       -6.769e-02  2.053e-01  -0.330  0.74167    
LOC_271                        2.924e-01  2.413e-01   1.211  0.22574    
LOC_281                       -1.208e-01  2.496e-01  -0.484  0.62833    
LOC_291                       -3.460e-01  1.748e-01  -1.980  0.04775 *  
LOC_301                        6.848e-01  2.441e-01   2.805  0.00503 ** 
LOC_311                        1.369e-01  3.374e-01   0.406  0.68483    
LOC_321                        2.357e-01  2.475e-01   0.953  0.34084    
LOC_331                        2.018e-01  5.201e-01   0.388  0.69807    
LOC_341                       -1.786e-01  1.993e-01  -0.896  0.37003    
LOC_351                        4.768e-01  1.580e-01   3.018  0.00254 ** 
LOC_361                       -2.524e-02  1.468e-01  -0.172  0.86350    
LOC_371                        7.174e-02  1.584e-01   0.453  0.65066    
LOC_381                        4.929e-01  3.690e-01   1.336  0.18169    
LOC_391                       -2.868e-01  1.489e-01  -1.927  0.05402 .  
LOC_401                        1.820e-01  1.711e-01   1.064  0.28748    
LOC_411                        1.194e-01  2.014e-01   0.593  0.55347    
LOC_421                       -2.913e-01  1.493e-01  -1.951  0.05104 .  
LOC_441                        3.424e-01  4.931e-01   0.694  0.48753    
LOC_451                        6.287e-02  2.286e-01   0.275  0.78327    
LOC_461                        8.187e-01  2.578e-01   3.175  0.00150 ** 
LOC_471                       -1.750e-01  2.141e-01  -0.817  0.41383    
LOC_481                        2.554e-01  1.028e-01   2.486  0.01293 *  
LOC_491                       -1.810e-01  2.129e-01  -0.850  0.39529    
LOC_501                       -4.051e-01  5.943e-01  -0.682  0.49548    
LOC_511                       -1.518e-01  8.935e-02  -1.699  0.08929 .  
LOC_531                        4.017e-01  1.355e-01   2.964  0.00304 ** 
LOC_541                       -5.801e-02  2.695e-01  -0.215  0.82958    
LOC_551                       -3.173e-01  2.539e-01  -1.250  0.21138    
TOA_151                        6.818e-02  6.912e-02   0.986  0.32391    
TOA_201                        1.128e+00  1.928e-01   5.850 4.93e-09 ***
TOA_301                        3.471e-01  8.905e-02   3.898 9.71e-05 ***
TOA_321                        4.200e-01  7.540e-01   0.557  0.57752    
TOA_351                       -5.349e-01  2.867e-01  -1.866  0.06204 .  
TOA_381                        3.283e-01  6.729e-02   4.879 1.07e-06 ***
TOA_401                       -1.052e-01  1.764e-01  -0.597  0.55079    
TOA_421                        1.141e+00  4.414e-01   2.585  0.00975 ** 
TOA_441                        2.166e+00  1.073e+00   2.020  0.04339 *  
PPGROUP_111                   -4.091e-01  1.172e-01  -3.492  0.00048 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 20683  on 14919  degrees of freedom
Residual deviance: 14543  on 14845  degrees of freedom
AIC: 14693

Number of Fisher Scoring iterations: 5

                                   VIF
GSEGRD                        1.687961
SalaryOverUnderIndAvg         1.359393
LowerLimitAge                 2.325664
BLS_FEDERAL_OtherSep_Rate     1.780297
BLS_FEDERAL_Quits_Rate        1.239183
BLS_FEDERAL_JobOpenings_Level 1.189397
BLS_FEDERAL_Layoffs_Rate      1.400682
LOSSqrt                       1.863295
SEPCount_EFDATE_OCCLog        1.108137
AGELVL_B                      1.168649
AGELVL_C                      1.661953
AGELVL_E                      1.407636
AGELVL_F                      1.284158
AGELVL_G                      1.270875
AGELVL_H                      1.360774
AGELVL_I                      1.516882
LOC_01                        1.137639
LOC_02                        1.069443
LOC_04                        1.263589
LOC_05                        1.077198
LOC_06                        1.572206
LOC_08                        1.247572
LOC_09                        1.038844
LOC_10                        1.013493
LOC_12                        1.294234
LOC_13                        1.305744
LOC_15                        1.152869
LOC_16                        1.063066
LOC_17                        1.171818
LOC_18                        1.065537
LOC_19                        1.041875
LOC_20                        1.066436
LOC_22                        1.093275
LOC_23                        1.037582
LOC_24                        1.587481
LOC_25                        1.103363
LOC_26                        1.100598
LOC_27                        1.073679
LOC_28                        1.067816
LOC_29                        1.147164
LOC_30                        1.081931
LOC_31                        1.039929
LOC_32                        1.071660
LOC_33                        1.017328
LOC_34                        1.096765
LOC_35                        1.208682
LOC_36                        1.187650
LOC_37                        1.175472
LOC_38                        1.034809
LOC_39                        1.192786
LOC_40                        1.147430
LOC_41                        1.108226
LOC_42                        1.180730
LOC_44                        1.017009
LOC_45                        1.082438
LOC_46                        1.078691
LOC_47                        1.092249
LOC_48                        1.525339
LOC_49                        1.091052
LOC_50                        1.013096
LOC_51                        1.693946
LOC_53                        1.246368
LOC_54                        1.053254
LOC_55                        1.069242
TOA_15                        1.405635
TOA_20                        1.028583
TOA_30                        1.243128
TOA_32                        1.005938
TOA_35                        1.077989
TOA_38                        1.176425
TOA_40                        1.062152
TOA_42                        1.010677
TOA_44                        1.004296
PPGROUP_11                    1.157491

Removed AFTER this step: LOC_09 


Call:
glm(formula = fla, family = "binomial", data = OPMAnalysisDataNoFamBinary)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-2.8257  -0.7391  -0.1369   0.8188   3.2362  

Coefficients:
                                Estimate Std. Error z value Pr(>|z|)    
(Intercept)                    2.156e+00  3.477e-01   6.202 5.58e-10 ***
GSEGRD                        -6.271e-02  1.531e-02  -4.096 4.21e-05 ***
SalaryOverUnderIndAvg          9.067e-07  2.405e-06   0.377 0.706118    
LowerLimitAge                 -4.495e-02  2.660e-03 -16.897  < 2e-16 ***
BLS_FEDERAL_OtherSep_Rate      3.240e+00  3.376e-01   9.599  < 2e-16 ***
BLS_FEDERAL_Quits_Rate         4.761e-01  3.158e-01   1.508 0.131633    
BLS_FEDERAL_JobOpenings_Level  1.706e-02  1.931e-03   8.839  < 2e-16 ***
BLS_FEDERAL_Layoffs_Rate       4.251e-02  1.230e-01   0.346 0.729666    
LOSSqrt                       -7.826e-01  2.293e-02 -34.132  < 2e-16 ***
SEPCount_EFDATE_OCCLog        -3.111e-02  1.610e-02  -1.932 0.053346 .  
AGELVL_B1                     -5.856e-01  2.200e-01  -2.662 0.007764 ** 
AGELVL_C1                     -1.677e-01  8.533e-02  -1.965 0.049429 *  
AGELVL_E1                      2.742e-01  6.732e-02   4.073 4.65e-05 ***
AGELVL_F1                      4.895e-01  7.015e-02   6.977 3.01e-12 ***
AGELVL_G1                      7.123e-01  7.399e-02   9.626  < 2e-16 ***
AGELVL_H1                      8.468e-01  7.623e-02  11.108  < 2e-16 ***
AGELVL_I1                      7.176e-01  8.111e-02   8.848  < 2e-16 ***
LOC_011                       -2.036e-01  1.728e-01  -1.178 0.238719    
LOC_021                        1.609e-01  2.546e-01   0.632 0.527466    
LOC_041                        5.336e-01  1.446e-01   3.690 0.000225 ***
LOC_051                        8.797e-02  2.370e-01   0.371 0.710477    
LOC_061                        4.165e-01  9.768e-02   4.264 2.01e-05 ***
LOC_081                        1.500e-01  1.324e-01   1.133 0.257194    
LOC_101                        2.032e-01  5.827e-01   0.349 0.727241    
LOC_121                        4.090e-02  1.267e-01   0.323 0.746827    
LOC_131                        1.756e-01  1.232e-01   1.426 0.153930    
LOC_151                       -6.258e-02  1.660e-01  -0.377 0.706187    
LOC_161                        5.888e-02  2.742e-01   0.215 0.829957    
LOC_171                       -2.158e-01  1.498e-01  -1.441 0.149706    
LOC_181                       -2.216e-01  2.541e-01  -0.872 0.382975    
LOC_191                        2.218e-01  3.221e-01   0.688 0.491157    
LOC_201                        6.199e-01  2.594e-01   2.390 0.016850 *  
LOC_221                       -1.054e-01  2.117e-01  -0.498 0.618403    
LOC_231                        7.513e-02  3.394e-01   0.221 0.824795    
LOC_241                       -1.039e-01  9.149e-02  -1.136 0.256143    
LOC_251                        1.494e-01  1.854e-01   0.806 0.420288    
LOC_261                       -6.571e-02  2.050e-01  -0.321 0.748543    
LOC_271                        2.944e-01  2.410e-01   1.222 0.221896    
LOC_281                       -1.189e-01  2.493e-01  -0.477 0.633497    
LOC_291                       -3.440e-01  1.744e-01  -1.973 0.048474 *  
LOC_301                        6.869e-01  2.438e-01   2.818 0.004838 ** 
LOC_311                        1.389e-01  3.372e-01   0.412 0.680409    
LOC_321                        2.377e-01  2.471e-01   0.962 0.336065    
LOC_331                        2.038e-01  5.199e-01   0.392 0.695063    
LOC_341                       -1.768e-01  1.989e-01  -0.889 0.374233    
LOC_351                        4.788e-01  1.575e-01   3.040 0.002367 ** 
LOC_361                       -2.333e-02  1.463e-01  -0.159 0.873316    
LOC_371                        7.368e-02  1.580e-01   0.466 0.640904    
LOC_381                        4.949e-01  3.688e-01   1.342 0.179617    
LOC_391                       -2.849e-01  1.484e-01  -1.920 0.054855 .  
LOC_401                        1.839e-01  1.707e-01   1.077 0.281257    
LOC_411                        1.213e-01  2.011e-01   0.603 0.546185    
LOC_421                       -2.895e-01  1.489e-01  -1.945 0.051828 .  
LOC_441                        3.444e-01  4.930e-01   0.699 0.484861    
LOC_451                        6.485e-02  2.282e-01   0.284 0.776292    
LOC_461                        8.208e-01  2.575e-01   3.187 0.001436 ** 
LOC_471                       -1.730e-01  2.138e-01  -0.809 0.418320    
LOC_481                        2.574e-01  1.021e-01   2.522 0.011663 *  
LOC_491                       -1.790e-01  2.125e-01  -0.842 0.399616    
LOC_501                       -4.031e-01  5.941e-01  -0.678 0.497491    
LOC_511                       -1.500e-01  8.861e-02  -1.693 0.090533 .  
LOC_531                        4.036e-01  1.350e-01   2.990 0.002788 ** 
LOC_541                       -5.610e-02  2.693e-01  -0.208 0.834962    
LOC_551                       -3.153e-01  2.536e-01  -1.243 0.213726    
TOA_151                        6.805e-02  6.911e-02   0.985 0.324764    
TOA_201                        1.128e+00  1.928e-01   5.851 4.88e-09 ***
TOA_301                        3.475e-01  8.900e-02   3.905 9.43e-05 ***
TOA_321                        4.197e-01  7.541e-01   0.557 0.577788    
TOA_351                       -5.343e-01  2.867e-01  -1.864 0.062320 .  
TOA_381                        3.280e-01  6.727e-02   4.876 1.08e-06 ***
TOA_401                       -1.051e-01  1.764e-01  -0.596 0.551374    
TOA_421                        1.141e+00  4.414e-01   2.585 0.009726 ** 
TOA_441                        2.168e+00  1.072e+00   2.021 0.043244 *  
PPGROUP_111                   -4.090e-01  1.172e-01  -3.491 0.000481 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 20683  on 14919  degrees of freedom
Residual deviance: 14543  on 14846  degrees of freedom
AIC: 14691

Number of Fisher Scoring iterations: 5

                                   VIF
GSEGRD                        1.682643
SalaryOverUnderIndAvg         1.358526
LowerLimitAge                 2.324361
BLS_FEDERAL_OtherSep_Rate     1.780344
BLS_FEDERAL_Quits_Rate        1.239175
BLS_FEDERAL_JobOpenings_Level 1.189204
BLS_FEDERAL_Layoffs_Rate      1.400701
LOSSqrt                       1.863038
SEPCount_EFDATE_OCCLog        1.108135
AGELVL_B                      1.168388
AGELVL_C                      1.661927
AGELVL_E                      1.407413
AGELVL_F                      1.284126
AGELVL_G                      1.270561
AGELVL_H                      1.360678
AGELVL_I                      1.516775
LOC_01                        1.132667
LOC_02                        1.066681
LOC_04                        1.254313
LOC_05                        1.074380
LOC_06                        1.547704
LOC_08                        1.237732
LOC_10                        1.012969
LOC_12                        1.283232
LOC_13                        1.294435
LOC_15                        1.146631
LOC_16                        1.060904
LOC_17                        1.164865
LOC_18                        1.063023
LOC_19                        1.040327
LOC_20                        1.064050
LOC_22                        1.089639
LOC_23                        1.036233
LOC_24                        1.565215
LOC_25                        1.099081
LOC_26                        1.096610
LOC_27                        1.070737
LOC_28                        1.065270
LOC_29                        1.141599
LOC_30                        1.078836
LOC_31                        1.038588
LOC_32                        1.068864
LOC_33                        1.016711
LOC_34                        1.093039
LOC_35                        1.201465
LOC_36                        1.179866
LOC_37                        1.168641
LOC_38                        1.033569
LOC_39                        1.185309
LOC_40                        1.141838
LOC_41                        1.104098
LOC_42                        1.173979
LOC_44                        1.016359
LOC_45                        1.079273
LOC_46                        1.075906
LOC_47                        1.088696
LOC_48                        1.504064
LOC_49                        1.087529
LOC_50                        1.012661
LOC_51                        1.665897
LOC_53                        1.236436
LOC_54                        1.051193
LOC_55                        1.066639
TOA_15                        1.405445
TOA_20                        1.028496
TOA_30                        1.241826
TOA_32                        1.005933
TOA_35                        1.077785
TOA_38                        1.175781
TOA_40                        1.062126
TOA_42                        1.010652
TOA_44                        1.004229
PPGROUP_11                    1.157450

Removed AFTER this step: LOC_36 


Call:
glm(formula = fla, family = "binomial", data = OPMAnalysisDataNoFamBinary)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-2.8256  -0.7391  -0.1369   0.8186   3.2371  

Coefficients:
                                Estimate Std. Error z value Pr(>|z|)    
(Intercept)                    2.149e+00  3.449e-01   6.231 4.62e-10 ***
GSEGRD                        -6.241e-02  1.519e-02  -4.107 4.01e-05 ***
SalaryOverUnderIndAvg          9.050e-07  2.405e-06   0.376 0.706622    
LowerLimitAge                 -4.496e-02  2.659e-03 -16.905  < 2e-16 ***
BLS_FEDERAL_OtherSep_Rate      3.240e+00  3.376e-01   9.598  < 2e-16 ***
BLS_FEDERAL_Quits_Rate         4.761e-01  3.158e-01   1.508 0.131609    
BLS_FEDERAL_JobOpenings_Level  1.707e-02  1.930e-03   8.841  < 2e-16 ***
BLS_FEDERAL_Layoffs_Rate       4.265e-02  1.230e-01   0.347 0.728828    
LOSSqrt                       -7.827e-01  2.292e-02 -34.146  < 2e-16 ***
SEPCount_EFDATE_OCCLog        -3.116e-02  1.610e-02  -1.936 0.052864 .  
AGELVL_B1                     -5.853e-01  2.200e-01  -2.661 0.007790 ** 
AGELVL_C1                     -1.677e-01  8.533e-02  -1.965 0.049440 *  
AGELVL_E1                      2.740e-01  6.731e-02   4.071 4.69e-05 ***
AGELVL_F1                      4.893e-01  7.015e-02   6.976 3.05e-12 ***
AGELVL_G1                      7.121e-01  7.399e-02   9.625  < 2e-16 ***
AGELVL_H1                      8.467e-01  7.623e-02  11.108  < 2e-16 ***
AGELVL_I1                      7.178e-01  8.110e-02   8.850  < 2e-16 ***
LOC_011                       -1.997e-01  1.711e-01  -1.167 0.243039    
LOC_021                        1.651e-01  2.533e-01   0.652 0.514460    
LOC_041                        5.378e-01  1.423e-01   3.780 0.000157 ***
LOC_051                        9.208e-02  2.356e-01   0.391 0.695898    
LOC_061                        4.205e-01  9.443e-02   4.453 8.47e-06 ***
LOC_081                        1.539e-01  1.300e-01   1.184 0.236349    
LOC_101                        2.075e-01  5.821e-01   0.356 0.721521    
LOC_121                        4.488e-02  1.242e-01   0.361 0.717851    
LOC_131                        1.795e-01  1.207e-01   1.487 0.136913    
LOC_151                       -5.857e-02  1.641e-01  -0.357 0.721125    
LOC_161                        6.303e-02  2.729e-01   0.231 0.817353    
LOC_171                       -2.119e-01  1.478e-01  -1.434 0.151637    
LOC_181                       -2.176e-01  2.528e-01  -0.861 0.389363    
LOC_191                        2.259e-01  3.211e-01   0.703 0.481749    
LOC_201                        6.240e-01  2.581e-01   2.418 0.015619 *  
LOC_221                       -1.014e-01  2.101e-01  -0.482 0.629481    
LOC_231                        7.918e-02  3.384e-01   0.234 0.814997    
LOC_241                       -1.003e-01  8.862e-02  -1.131 0.257852    
LOC_251                        1.533e-01  1.838e-01   0.834 0.404307    
LOC_261                       -6.157e-02  2.033e-01  -0.303 0.762028    
LOC_271                        2.986e-01  2.396e-01   1.247 0.212559    
LOC_281                       -1.148e-01  2.480e-01  -0.463 0.643378    
LOC_291                       -3.400e-01  1.725e-01  -1.971 0.048702 *  
LOC_301                        6.911e-01  2.423e-01   2.852 0.004340 ** 
LOC_311                        1.430e-01  3.362e-01   0.425 0.670499    
LOC_321                        2.419e-01  2.458e-01   0.984 0.325121    
LOC_331                        2.081e-01  5.193e-01   0.401 0.688614    
LOC_341                       -1.729e-01  1.974e-01  -0.876 0.381219    
LOC_351                        4.829e-01  1.553e-01   3.109 0.001878 ** 
LOC_371                        7.775e-02  1.559e-01   0.499 0.617944    
LOC_381                        4.991e-01  3.678e-01   1.357 0.174791    
LOC_391                       -2.809e-01  1.463e-01  -1.921 0.054767 .  
LOC_401                        1.879e-01  1.688e-01   1.113 0.265539    
LOC_411                        1.254e-01  1.994e-01   0.629 0.529422    
LOC_421                       -2.856e-01  1.468e-01  -1.945 0.051802 .  
LOC_441                        3.484e-01  4.924e-01   0.708 0.479219    
LOC_451                        6.898e-02  2.267e-01   0.304 0.760963    
LOC_461                        8.252e-01  2.560e-01   3.223 0.001269 ** 
LOC_471                       -1.690e-01  2.123e-01  -0.796 0.426036    
LOC_481                        2.614e-01  9.888e-02   2.644 0.008201 ** 
LOC_491                       -1.750e-01  2.110e-01  -0.829 0.406951    
LOC_501                       -3.991e-01  5.936e-01  -0.672 0.501407    
LOC_511                       -1.462e-01  8.531e-02  -1.713 0.086653 .  
LOC_531                        4.077e-01  1.326e-01   3.075 0.002108 ** 
LOC_541                       -5.214e-02  2.681e-01  -0.194 0.845804    
LOC_551                       -3.111e-01  2.522e-01  -1.234 0.217366    
TOA_151                        6.799e-02  6.911e-02   0.984 0.325235    
TOA_201                        1.128e+00  1.928e-01   5.852 4.85e-09 ***
TOA_301                        3.481e-01  8.894e-02   3.913 9.11e-05 ***
TOA_321                        4.194e-01  7.541e-01   0.556 0.578107    
TOA_351                       -5.338e-01  2.866e-01  -1.862 0.062570 .  
TOA_381                        3.276e-01  6.721e-02   4.874 1.09e-06 ***
TOA_401                       -1.044e-01  1.763e-01  -0.592 0.553758    
TOA_421                        1.141e+00  4.414e-01   2.586 0.009714 ** 
TOA_441                        2.171e+00  1.072e+00   2.024 0.042924 *  
PPGROUP_111                   -4.087e-01  1.171e-01  -3.488 0.000486 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 20683  on 14919  degrees of freedom
Residual deviance: 14544  on 14847  degrees of freedom
AIC: 14690

Number of Fisher Scoring iterations: 5

                                   VIF
GSEGRD                        1.657660
SalaryOverUnderIndAvg         1.358527
LowerLimitAge                 2.323183
BLS_FEDERAL_OtherSep_Rate     1.780218
BLS_FEDERAL_Quits_Rate        1.239155
BLS_FEDERAL_JobOpenings_Level 1.189094
BLS_FEDERAL_Layoffs_Rate      1.400661
LOSSqrt                       1.861913
SEPCount_EFDATE_OCCLog        1.107623
AGELVL_B                      1.168345
AGELVL_C                      1.661945
AGELVL_E                      1.407111
AGELVL_F                      1.283799
AGELVL_G                      1.270373
AGELVL_H                      1.360653
AGELVL_I                      1.516606
LOC_01                        1.109664
LOC_02                        1.055243
LOC_04                        1.213554
LOC_05                        1.061690
LOC_06                        1.446196
LOC_08                        1.193965
LOC_10                        1.010876
LOC_12                        1.233387
LOC_13                        1.243048
LOC_15                        1.120307
LOC_16                        1.051323
LOC_17                        1.133941
LOC_18                        1.052290
LOC_19                        1.033660
LOC_20                        1.053602
LOC_22                        1.073896
LOC_23                        1.030415
LOC_24                        1.468350
LOC_25                        1.080265
LOC_26                        1.079018
LOC_27                        1.057836
LOC_28                        1.054198
LOC_29                        1.116958
LOC_30                        1.065860
LOC_31                        1.032410
LOC_32                        1.057221
LOC_33                        1.014011
LOC_34                        1.076617
LOC_35                        1.168674
LOC_37                        1.138096
LOC_38                        1.028209
LOC_39                        1.151499
LOC_40                        1.116769
LOC_41                        1.086205
LOC_42                        1.142333
LOC_44                        1.013695
LOC_45                        1.065360
LOC_46                        1.063668
LOC_47                        1.073126
LOC_48                        1.412089
LOC_49                        1.072259
LOC_50                        1.010826
LOC_51                        1.544234
LOC_53                        1.193008
LOC_54                        1.042254
LOC_55                        1.055436
TOA_15                        1.405458
TOA_20                        1.028455
TOA_30                        1.240230
TOA_32                        1.005925
TOA_35                        1.077661
TOA_38                        1.173642
TOA_40                        1.061551
TOA_42                        1.010642
TOA_44                        1.003920
PPGROUP_11                    1.156931

Removed AFTER this step: LOC_54 


Call:
glm(formula = fla, family = "binomial", data = OPMAnalysisDataNoFamBinary)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-2.8257  -0.7396  -0.1369   0.8188   3.2376  

Coefficients:
                                Estimate Std. Error z value Pr(>|z|)    
(Intercept)                    2.147e+00  3.446e-01   6.229 4.70e-10 ***
GSEGRD                        -6.229e-02  1.518e-02  -4.103 4.09e-05 ***
SalaryOverUnderIndAvg          9.247e-07  2.402e-06   0.385 0.700315    
LowerLimitAge                 -4.496e-02  2.659e-03 -16.911  < 2e-16 ***
BLS_FEDERAL_OtherSep_Rate      3.241e+00  3.375e-01   9.601  < 2e-16 ***
BLS_FEDERAL_Quits_Rate         4.753e-01  3.157e-01   1.505 0.132269    
BLS_FEDERAL_JobOpenings_Level  1.707e-02  1.930e-03   8.841  < 2e-16 ***
BLS_FEDERAL_Layoffs_Rate       4.268e-02  1.230e-01   0.347 0.728648    
LOSSqrt                       -7.828e-01  2.292e-02 -34.152  < 2e-16 ***
SEPCount_EFDATE_OCCLog        -3.120e-02  1.610e-02  -1.938 0.052605 .  
AGELVL_B1                     -5.848e-01  2.199e-01  -2.659 0.007834 ** 
AGELVL_C1                     -1.674e-01  8.532e-02  -1.962 0.049707 *  
AGELVL_E1                      2.741e-01  6.731e-02   4.072 4.66e-05 ***
AGELVL_F1                      4.893e-01  7.015e-02   6.976 3.03e-12 ***
AGELVL_G1                      7.122e-01  7.398e-02   9.626  < 2e-16 ***
AGELVL_H1                      8.468e-01  7.623e-02  11.109  < 2e-16 ***
AGELVL_I1                      7.177e-01  8.110e-02   8.850  < 2e-16 ***
LOC_011                       -1.973e-01  1.706e-01  -1.157 0.247438    
LOC_021                        1.674e-01  2.530e-01   0.662 0.508226    
LOC_041                        5.403e-01  1.417e-01   3.813 0.000137 ***
LOC_051                        9.451e-02  2.352e-01   0.402 0.687858    
LOC_061                        4.226e-01  9.377e-02   4.507 6.56e-06 ***
LOC_081                        1.562e-01  1.295e-01   1.207 0.227592    
LOC_101                        2.100e-01  5.819e-01   0.361 0.718237    
LOC_121                        4.725e-02  1.236e-01   0.382 0.702268    
LOC_131                        1.819e-01  1.201e-01   1.514 0.130008    
LOC_151                       -5.631e-02  1.637e-01  -0.344 0.730806    
LOC_161                        6.548e-02  2.726e-01   0.240 0.810172    
LOC_171                       -2.097e-01  1.474e-01  -1.423 0.154724    
LOC_181                       -2.152e-01  2.525e-01  -0.852 0.394088    
LOC_191                        2.284e-01  3.209e-01   0.712 0.476621    
LOC_201                        6.265e-01  2.578e-01   2.430 0.015092 *  
LOC_221                       -9.900e-02  2.098e-01  -0.472 0.636978    
LOC_231                        8.146e-02  3.382e-01   0.241 0.809668    
LOC_241                       -9.819e-02  8.797e-02  -1.116 0.264344    
LOC_251                        1.554e-01  1.835e-01   0.847 0.397018    
LOC_261                       -5.924e-02  2.030e-01  -0.292 0.770395    
LOC_271                        3.010e-01  2.392e-01   1.258 0.208339    
LOC_281                       -1.124e-01  2.477e-01  -0.454 0.650014    
LOC_291                       -3.375e-01  1.720e-01  -1.962 0.049728 *  
LOC_301                        6.936e-01  2.420e-01   2.866 0.004156 ** 
LOC_311                        1.455e-01  3.359e-01   0.433 0.664979    
LOC_321                        2.443e-01  2.455e-01   0.995 0.319647    
LOC_331                        2.106e-01  5.191e-01   0.406 0.684997    
LOC_341                       -1.707e-01  1.971e-01  -0.866 0.386370    
LOC_351                        4.854e-01  1.548e-01   3.135 0.001721 ** 
LOC_371                        8.013e-02  1.554e-01   0.516 0.606119    
LOC_381                        5.016e-01  3.676e-01   1.364 0.172440    
LOC_391                       -2.786e-01  1.458e-01  -1.911 0.055976 .  
LOC_401                        1.903e-01  1.684e-01   1.130 0.258468    
LOC_411                        1.278e-01  1.991e-01   0.642 0.520836    
LOC_421                       -2.833e-01  1.464e-01  -1.935 0.052931 .  
LOC_441                        3.507e-01  4.923e-01   0.712 0.476221    
LOC_451                        7.136e-02  2.264e-01   0.315 0.752626    
LOC_461                        8.277e-01  2.557e-01   3.237 0.001209 ** 
LOC_471                       -1.665e-01  2.119e-01  -0.786 0.431883    
LOC_481                        2.637e-01  9.816e-02   2.687 0.007212 ** 
LOC_491                       -1.726e-01  2.107e-01  -0.819 0.412554    
LOC_501                       -3.965e-01  5.934e-01  -0.668 0.504015    
LOC_511                       -1.440e-01  8.457e-02  -1.703 0.088650 .  
LOC_531                        4.099e-01  1.321e-01   3.104 0.001911 ** 
LOC_551                       -3.087e-01  2.519e-01  -1.225 0.220422    
TOA_151                        6.799e-02  6.911e-02   0.984 0.325227    
TOA_201                        1.129e+00  1.928e-01   5.855 4.78e-09 ***
TOA_301                        3.489e-01  8.883e-02   3.928 8.55e-05 ***
TOA_321                        4.193e-01  7.541e-01   0.556 0.578187    
TOA_351                       -5.332e-01  2.866e-01  -1.860 0.062829 .  
TOA_381                        3.272e-01  6.718e-02   4.870 1.11e-06 ***
TOA_401                       -1.038e-01  1.763e-01  -0.589 0.556003    
TOA_421                        1.140e+00  4.414e-01   2.584 0.009771 ** 
TOA_441                        2.173e+00  1.072e+00   2.026 0.042731 *  
PPGROUP_111                   -4.088e-01  1.171e-01  -3.490 0.000483 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 20683  on 14919  degrees of freedom
Residual deviance: 14544  on 14848  degrees of freedom
AIC: 14688

Number of Fisher Scoring iterations: 5

                                   VIF
GSEGRD                        1.654900
SalaryOverUnderIndAvg         1.356144
LowerLimitAge                 2.322537
BLS_FEDERAL_OtherSep_Rate     1.779967
BLS_FEDERAL_Quits_Rate        1.238896
BLS_FEDERAL_JobOpenings_Level 1.189080
BLS_FEDERAL_Layoffs_Rate      1.400649
LOSSqrt                       1.861550
SEPCount_EFDATE_OCCLog        1.107501
AGELVL_B                      1.168203
AGELVL_C                      1.661700
AGELVL_E                      1.407073
AGELVL_F                      1.283799
AGELVL_G                      1.270386
AGELVL_H                      1.360623
AGELVL_I                      1.516608
LOC_01                        1.104134
LOC_02                        1.052987
LOC_04                        1.203869
LOC_05                        1.058690
LOC_06                        1.426053
LOC_08                        1.184262
LOC_10                        1.010384
LOC_12                        1.221491
LOC_13                        1.230847
LOC_15                        1.114678
LOC_16                        1.049069
LOC_17                        1.127233
LOC_18                        1.049736
LOC_19                        1.032040
LOC_20                        1.051063
LOC_22                        1.070241
LOC_23                        1.029175
LOC_24                        1.446867
LOC_25                        1.076466
LOC_26                        1.075257
LOC_27                        1.055076
LOC_28                        1.051507
LOC_29                        1.111154
LOC_30                        1.063039
LOC_31                        1.030958
LOC_32                        1.054476
LOC_33                        1.013394
LOC_34                        1.073303
LOC_35                        1.161118
LOC_37                        1.131070
LOC_38                        1.027011
LOC_39                        1.143769
LOC_40                        1.111134
LOC_41                        1.082089
LOC_42                        1.135238
LOC_44                        1.013110
LOC_45                        1.062247
LOC_46                        1.060953
LOC_47                        1.069456
LOC_48                        1.391448
LOC_49                        1.068675
LOC_50                        1.010342
LOC_51                        1.517656
LOC_53                        1.183730
LOC_55                        1.052816
TOA_15                        1.405450
TOA_20                        1.028352
TOA_30                        1.236945
TOA_32                        1.005925
TOA_35                        1.077547
TOA_38                        1.172736
TOA_40                        1.061219
TOA_42                        1.010531
TOA_44                        1.003834
PPGROUP_11                    1.156872

Removed AFTER this step: LOC_16 


Call:
glm(formula = fla, family = "binomial", data = OPMAnalysisDataNoFamBinary)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-2.8259  -0.7395  -0.1369   0.8187   3.2369  

Coefficients:
                                Estimate Std. Error z value Pr(>|z|)    
(Intercept)                    2.154e+00  3.434e-01   6.271 3.58e-10 ***
GSEGRD                        -6.265e-02  1.511e-02  -4.146 3.38e-05 ***
SalaryOverUnderIndAvg          8.918e-07  2.398e-06   0.372 0.710021    
LowerLimitAge                 -4.495e-02  2.659e-03 -16.910  < 2e-16 ***
BLS_FEDERAL_OtherSep_Rate      3.239e+00  3.375e-01   9.599  < 2e-16 ***
BLS_FEDERAL_Quits_Rate         4.749e-01  3.157e-01   1.504 0.132570    
BLS_FEDERAL_JobOpenings_Level  1.707e-02  1.930e-03   8.844  < 2e-16 ***
BLS_FEDERAL_Layoffs_Rate       4.307e-02  1.230e-01   0.350 0.726209    
LOSSqrt                       -7.825e-01  2.289e-02 -34.183  < 2e-16 ***
SEPCount_EFDATE_OCCLog        -3.127e-02  1.609e-02  -1.943 0.051975 .  
AGELVL_B1                     -5.852e-01  2.199e-01  -2.661 0.007790 ** 
AGELVL_C1                     -1.674e-01  8.532e-02  -1.962 0.049730 *  
AGELVL_E1                      2.740e-01  6.731e-02   4.072 4.67e-05 ***
AGELVL_F1                      4.896e-01  7.013e-02   6.982 2.92e-12 ***
AGELVL_G1                      7.123e-01  7.399e-02   9.627  < 2e-16 ***
AGELVL_H1                      8.470e-01  7.623e-02  11.111  < 2e-16 ***
AGELVL_I1                      7.179e-01  8.110e-02   8.853  < 2e-16 ***
LOC_011                       -2.003e-01  1.702e-01  -1.177 0.239318    
LOC_021                        1.644e-01  2.527e-01   0.651 0.515262    
LOC_041                        5.371e-01  1.411e-01   3.807 0.000141 ***
LOC_051                        9.143e-02  2.349e-01   0.389 0.697093    
LOC_061                        4.200e-01  9.312e-02   4.510 6.48e-06 ***
LOC_081                        1.534e-01  1.289e-01   1.190 0.234154    
LOC_101                        2.069e-01  5.817e-01   0.356 0.722062    
LOC_121                        4.430e-02  1.230e-01   0.360 0.718710    
LOC_131                        1.790e-01  1.195e-01   1.498 0.134259    
LOC_151                       -5.906e-02  1.633e-01  -0.362 0.717530    
LOC_171                       -2.124e-01  1.470e-01  -1.446 0.148290    
LOC_181                       -2.180e-01  2.522e-01  -0.865 0.387234    
LOC_191                        2.252e-01  3.206e-01   0.703 0.482337    
LOC_201                        6.233e-01  2.574e-01   2.421 0.015467 *  
LOC_221                       -1.020e-01  2.094e-01  -0.487 0.626127    
LOC_231                        7.852e-02  3.380e-01   0.232 0.816269    
LOC_241                       -1.006e-01  8.737e-02  -1.152 0.249362    
LOC_251                        1.528e-01  1.831e-01   0.834 0.404166    
LOC_261                       -6.211e-02  2.026e-01  -0.307 0.759177    
LOC_271                        2.980e-01  2.389e-01   1.247 0.212309    
LOC_281                       -1.155e-01  2.473e-01  -0.467 0.640603    
LOC_291                       -3.406e-01  1.715e-01  -1.986 0.047075 *  
LOC_301                        6.903e-01  2.416e-01   2.857 0.004273 ** 
LOC_311                        1.423e-01  3.356e-01   0.424 0.671648    
LOC_321                        2.413e-01  2.452e-01   0.984 0.324962    
LOC_331                        2.076e-01  5.189e-01   0.400 0.689157    
LOC_341                       -1.734e-01  1.968e-01  -0.881 0.378371    
LOC_351                        4.822e-01  1.543e-01   3.126 0.001775 ** 
LOC_371                        7.726e-02  1.549e-01   0.499 0.618043    
LOC_381                        4.983e-01  3.673e-01   1.357 0.174921    
LOC_391                       -2.814e-01  1.453e-01  -1.937 0.052698 .  
LOC_401                        1.873e-01  1.679e-01   1.115 0.264686    
LOC_411                        1.247e-01  1.986e-01   0.628 0.530157    
LOC_421                       -2.860e-01  1.459e-01  -1.960 0.050015 .  
LOC_441                        3.478e-01  4.921e-01   0.707 0.479689    
LOC_451                        6.833e-02  2.261e-01   0.302 0.762458    
LOC_461                        8.244e-01  2.553e-01   3.229 0.001244 ** 
LOC_471                       -1.695e-01  2.115e-01  -0.801 0.422874    
LOC_481                        2.609e-01  9.743e-02   2.678 0.007416 ** 
LOC_491                       -1.756e-01  2.103e-01  -0.835 0.403642    
LOC_501                       -3.994e-01  5.933e-01  -0.673 0.500805    
LOC_511                       -1.467e-01  8.385e-02  -1.749 0.080279 .  
LOC_531                        4.070e-01  1.315e-01   3.095 0.001970 ** 
LOC_551                       -3.117e-01  2.516e-01  -1.239 0.215432    
TOA_151                        6.772e-02  6.910e-02   0.980 0.327049    
TOA_201                        1.128e+00  1.928e-01   5.853 4.83e-09 ***
TOA_301                        3.483e-01  8.878e-02   3.923 8.75e-05 ***
TOA_321                        4.197e-01  7.541e-01   0.557 0.577802    
TOA_351                       -5.345e-01  2.866e-01  -1.865 0.062149 .  
TOA_381                        3.272e-01  6.718e-02   4.870 1.12e-06 ***
TOA_401                       -1.043e-01  1.763e-01  -0.591 0.554262    
TOA_421                        1.140e+00  4.414e-01   2.582 0.009810 ** 
TOA_441                        2.171e+00  1.072e+00   2.025 0.042908 *  
PPGROUP_111                   -4.095e-01  1.171e-01  -3.496 0.000471 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 20683  on 14919  degrees of freedom
Residual deviance: 14544  on 14849  degrees of freedom
AIC: 14686

Number of Fisher Scoring iterations: 5

                                   VIF
GSEGRD                        1.638746
SalaryOverUnderIndAvg         1.351672
LowerLimitAge                 2.321901
BLS_FEDERAL_OtherSep_Rate     1.779528
BLS_FEDERAL_Quits_Rate        1.238885
BLS_FEDERAL_JobOpenings_Level 1.188952
BLS_FEDERAL_Layoffs_Rate      1.400393
LOSSqrt                       1.856960
SEPCount_EFDATE_OCCLog        1.107038
AGELVL_B                      1.168157
AGELVL_C                      1.661675
AGELVL_E                      1.407045
AGELVL_F                      1.283355
AGELVL_G                      1.270317
AGELVL_H                      1.360518
AGELVL_I                      1.516401
LOC_01                        1.098556
LOC_02                        1.050485
LOC_04                        1.193535
LOC_05                        1.055547
LOC_06                        1.406677
LOC_08                        1.174566
LOC_10                        1.009912
LOC_12                        1.209466
LOC_13                        1.218387
LOC_15                        1.109244
LOC_17                        1.120707
LOC_18                        1.047377
LOC_19                        1.030321
LOC_20                        1.048376
LOC_22                        1.066426
LOC_23                        1.027841
LOC_24                        1.427402
LOC_25                        1.072694
LOC_26                        1.071532
LOC_27                        1.052160
LOC_28                        1.048686
LOC_29                        1.105063
LOC_30                        1.059740
LOC_31                        1.029346
LOC_32                        1.051781
LOC_33                        1.012808
LOC_34                        1.070013
LOC_35                        1.152912
LOC_37                        1.124396
LOC_38                        1.025631
LOC_39                        1.136152
LOC_40                        1.105113
LOC_41                        1.077513
LOC_42                        1.128536
LOC_44                        1.012508
LOC_45                        1.058950
LOC_46                        1.057952
LOC_47                        1.065801
LOC_48                        1.370902
LOC_49                        1.064935
LOC_50                        1.009927
LOC_51                        1.491672
LOC_53                        1.173919
LOC_55                        1.050285
TOA_15                        1.405102
TOA_20                        1.028264
TOA_30                        1.235620
TOA_32                        1.005920
TOA_35                        1.077184
TOA_38                        1.172716
TOA_40                        1.061095
TOA_42                        1.010499
TOA_44                        1.003777
PPGROUP_11                    1.156240

Removed AFTER this step: LOC_23 


Call:
glm(formula = fla, family = "binomial", data = OPMAnalysisDataNoFamBinary)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-2.8260  -0.7397  -0.1369   0.8187   3.2365  

Coefficients:
                                Estimate Std. Error z value Pr(>|z|)    
(Intercept)                    2.158e+00  3.428e-01   6.295 3.07e-10 ***
GSEGRD                        -6.280e-02  1.509e-02  -4.161 3.17e-05 ***
SalaryOverUnderIndAvg          8.910e-07  2.398e-06   0.372 0.710264    
LowerLimitAge                 -4.495e-02  2.659e-03 -16.909  < 2e-16 ***
BLS_FEDERAL_OtherSep_Rate      3.239e+00  3.375e-01   9.597  < 2e-16 ***
BLS_FEDERAL_Quits_Rate         4.754e-01  3.157e-01   1.506 0.132148    
BLS_FEDERAL_JobOpenings_Level  1.707e-02  1.930e-03   8.842  < 2e-16 ***
BLS_FEDERAL_Layoffs_Rate       4.286e-02  1.230e-01   0.348 0.727502    
LOSSqrt                       -7.824e-01  2.289e-02 -34.184  < 2e-16 ***
SEPCount_EFDATE_OCCLog        -3.132e-02  1.609e-02  -1.947 0.051586 .  
AGELVL_B1                     -5.830e-01  2.197e-01  -2.654 0.007966 ** 
AGELVL_C1                     -1.673e-01  8.532e-02  -1.961 0.049880 *  
AGELVL_E1                      2.739e-01  6.731e-02   4.069 4.71e-05 ***
AGELVL_F1                      4.896e-01  7.013e-02   6.980 2.94e-12 ***
AGELVL_G1                      7.123e-01  7.398e-02   9.628  < 2e-16 ***
AGELVL_H1                      8.470e-01  7.622e-02  11.112  < 2e-16 ***
AGELVL_I1                      7.182e-01  8.109e-02   8.856  < 2e-16 ***
LOC_011                       -2.023e-01  1.700e-01  -1.190 0.233961    
LOC_021                        1.623e-01  2.525e-01   0.643 0.520500    
LOC_041                        5.350e-01  1.408e-01   3.800 0.000145 ***
LOC_051                        8.937e-02  2.347e-01   0.381 0.703395    
LOC_061                        4.180e-01  9.273e-02   4.508 6.55e-06 ***
LOC_081                        1.514e-01  1.287e-01   1.177 0.239263    
LOC_101                        2.048e-01  5.817e-01   0.352 0.724711    
LOC_121                        4.229e-02  1.227e-01   0.345 0.730350    
LOC_131                        1.770e-01  1.192e-01   1.485 0.137625    
LOC_151                       -6.109e-02  1.630e-01  -0.375 0.707854    
LOC_171                       -2.144e-01  1.467e-01  -1.462 0.143862    
LOC_181                       -2.200e-01  2.520e-01  -0.873 0.382669    
LOC_191                        2.231e-01  3.204e-01   0.696 0.486294    
LOC_201                        6.212e-01  2.573e-01   2.415 0.015753 *  
LOC_221                       -1.041e-01  2.092e-01  -0.497 0.618910    
LOC_241                       -1.025e-01  8.701e-02  -1.178 0.238768    
LOC_251                        1.508e-01  1.830e-01   0.824 0.409700    
LOC_261                       -6.416e-02  2.024e-01  -0.317 0.751252    
LOC_271                        2.958e-01  2.387e-01   1.239 0.215273    
LOC_281                       -1.175e-01  2.472e-01  -0.476 0.634400    
LOC_291                       -3.427e-01  1.713e-01  -2.000 0.045464 *  
LOC_301                        6.881e-01  2.414e-01   2.850 0.004366 ** 
LOC_311                        1.401e-01  3.355e-01   0.418 0.676204    
LOC_321                        2.393e-01  2.450e-01   0.977 0.328762    
LOC_331                        2.054e-01  5.188e-01   0.396 0.692259    
LOC_341                       -1.754e-01  1.966e-01  -0.892 0.372279    
LOC_351                        4.801e-01  1.540e-01   3.117 0.001825 ** 
LOC_371                        7.524e-02  1.547e-01   0.486 0.626693    
LOC_381                        4.961e-01  3.672e-01   1.351 0.176688    
LOC_391                       -2.835e-01  1.450e-01  -1.955 0.050596 .  
LOC_401                        1.852e-01  1.677e-01   1.105 0.269263    
LOC_411                        1.226e-01  1.984e-01   0.618 0.536760    
LOC_421                       -2.880e-01  1.457e-01  -1.977 0.048035 *  
LOC_441                        3.457e-01  4.920e-01   0.703 0.482286    
LOC_451                        6.629e-02  2.259e-01   0.293 0.769168    
LOC_461                        8.223e-01  2.552e-01   3.222 0.001271 ** 
LOC_471                       -1.716e-01  2.113e-01  -0.812 0.416850    
LOC_481                        2.589e-01  9.705e-02   2.667 0.007643 ** 
LOC_491                       -1.777e-01  2.101e-01  -0.846 0.397728    
LOC_501                       -4.015e-01  5.932e-01  -0.677 0.498578    
LOC_511                       -1.486e-01  8.341e-02  -1.782 0.074736 .  
LOC_531                        4.049e-01  1.312e-01   3.086 0.002028 ** 
LOC_551                       -3.137e-01  2.514e-01  -1.248 0.212070    
TOA_151                        6.773e-02  6.910e-02   0.980 0.327002    
TOA_201                        1.128e+00  1.928e-01   5.851 4.89e-09 ***
TOA_301                        3.476e-01  8.873e-02   3.917 8.95e-05 ***
TOA_321                        4.198e-01  7.540e-01   0.557 0.577705    
TOA_351                       -5.359e-01  2.865e-01  -1.870 0.061420 .  
TOA_381                        3.270e-01  6.718e-02   4.868 1.13e-06 ***
TOA_401                       -1.048e-01  1.763e-01  -0.595 0.552046    
TOA_421                        1.139e+00  4.414e-01   2.581 0.009841 ** 
TOA_441                        2.169e+00  1.072e+00   2.023 0.043047 *  
PPGROUP_111                   -4.096e-01  1.171e-01  -3.498 0.000469 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 20683  on 14919  degrees of freedom
Residual deviance: 14544  on 14850  degrees of freedom
AIC: 14684

Number of Fisher Scoring iterations: 5

                                   VIF
GSEGRD                        1.635493
SalaryOverUnderIndAvg         1.351637
LowerLimitAge                 2.321890
BLS_FEDERAL_OtherSep_Rate     1.779447
BLS_FEDERAL_Quits_Rate        1.238832
BLS_FEDERAL_JobOpenings_Level 1.188706
BLS_FEDERAL_Layoffs_Rate      1.400361
LOSSqrt                       1.856547
SEPCount_EFDATE_OCCLog        1.106852
AGELVL_B                      1.166101
AGELVL_C                      1.661686
AGELVL_E                      1.406940
AGELVL_F                      1.283334
AGELVL_G                      1.270330
AGELVL_H                      1.360533
AGELVL_I                      1.516174
LOC_01                        1.095654
LOC_02                        1.049075
LOC_04                        1.188659
LOC_05                        1.054045
LOC_06                        1.394802
LOC_08                        1.169349
LOC_10                        1.009673
LOC_12                        1.203474
LOC_13                        1.212179
LOC_15                        1.106082
LOC_17                        1.116947
LOC_18                        1.046191
LOC_19                        1.029487
LOC_20                        1.047073
LOC_22                        1.064552
LOC_24                        1.415517
LOC_25                        1.070446
LOC_26                        1.069504
LOC_27                        1.050601
LOC_28                        1.047330
LOC_29                        1.102109
LOC_30                        1.058144
LOC_31                        1.028569
LOC_32                        1.050431
LOC_33                        1.012465
LOC_34                        1.067863
LOC_35                        1.149027
LOC_37                        1.120893
LOC_38                        1.024947
LOC_39                        1.132050
LOC_40                        1.102110
LOC_41                        1.075232
LOC_42                        1.124592
LOC_44                        1.012161
LOC_45                        1.057364
LOC_46                        1.056550
LOC_47                        1.063935
LOC_48                        1.360220
LOC_49                        1.063052
LOC_50                        1.009711
LOC_51                        1.476127
LOC_53                        1.168301
LOC_55                        1.048946
TOA_15                        1.405138
TOA_20                        1.028186
TOA_30                        1.234263
TOA_32                        1.005920
TOA_35                        1.076690
TOA_38                        1.172635
TOA_40                        1.060882
TOA_42                        1.010477
TOA_44                        1.003742
PPGROUP_11                    1.156210

Removed AFTER this step: LOC_45 


Call:
glm(formula = fla, family = "binomial", data = OPMAnalysisDataNoFamBinary)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-2.8264  -0.7398  -0.1369   0.8186   3.2359  

Coefficients:
                                Estimate Std. Error z value Pr(>|z|)    
(Intercept)                    2.167e+00  3.416e-01   6.344 2.24e-10 ***
GSEGRD                        -6.319e-02  1.504e-02  -4.203 2.63e-05 ***
SalaryOverUnderIndAvg          8.598e-07  2.396e-06   0.359 0.719708    
LowerLimitAge                 -4.493e-02  2.657e-03 -16.909  < 2e-16 ***
BLS_FEDERAL_OtherSep_Rate      3.239e+00  3.375e-01   9.598  < 2e-16 ***
BLS_FEDERAL_Quits_Rate         4.745e-01  3.157e-01   1.503 0.132832    
BLS_FEDERAL_JobOpenings_Level  1.706e-02  1.930e-03   8.840  < 2e-16 ***
BLS_FEDERAL_Layoffs_Rate       4.296e-02  1.230e-01   0.349 0.726862    
LOSSqrt                       -7.824e-01  2.289e-02 -34.185  < 2e-16 ***
SEPCount_EFDATE_OCCLog        -3.135e-02  1.609e-02  -1.948 0.051365 .  
AGELVL_B1                     -5.840e-01  2.197e-01  -2.658 0.007851 ** 
AGELVL_C1                     -1.674e-01  8.532e-02  -1.962 0.049720 *  
AGELVL_E1                      2.739e-01  6.731e-02   4.069 4.73e-05 ***
AGELVL_F1                      4.899e-01  7.012e-02   6.986 2.82e-12 ***
AGELVL_G1                      7.129e-01  7.396e-02   9.639  < 2e-16 ***
AGELVL_H1                      8.474e-01  7.621e-02  11.120  < 2e-16 ***
AGELVL_I1                      7.184e-01  8.109e-02   8.860  < 2e-16 ***
LOC_011                       -2.061e-01  1.695e-01  -1.216 0.223980    
LOC_021                        1.583e-01  2.522e-01   0.628 0.530074    
LOC_041                        5.308e-01  1.401e-01   3.790 0.000151 ***
LOC_051                        8.535e-02  2.343e-01   0.364 0.715696    
LOC_061                        4.144e-01  9.193e-02   4.508 6.53e-06 ***
LOC_081                        1.476e-01  1.280e-01   1.153 0.248811    
LOC_101                        2.008e-01  5.815e-01   0.345 0.729929    
LOC_121                        3.837e-02  1.220e-01   0.315 0.753075    
LOC_131                        1.731e-01  1.185e-01   1.461 0.143930    
LOC_151                       -6.485e-02  1.625e-01  -0.399 0.689882    
LOC_171                       -2.180e-01  1.462e-01  -1.492 0.135829    
LOC_181                       -2.239e-01  2.517e-01  -0.889 0.373782    
LOC_191                        2.190e-01  3.201e-01   0.684 0.493838    
LOC_201                        6.172e-01  2.569e-01   2.402 0.016298 *  
LOC_221                       -1.080e-01  2.088e-01  -0.517 0.604829    
LOC_241                       -1.058e-01  8.626e-02  -1.227 0.219857    
LOC_251                        1.473e-01  1.826e-01   0.807 0.419696    
LOC_261                       -6.805e-02  2.020e-01  -0.337 0.736181    
LOC_271                        2.918e-01  2.383e-01   1.224 0.220796    
LOC_281                       -1.216e-01  2.468e-01  -0.493 0.622356    
LOC_291                       -3.467e-01  1.708e-01  -2.030 0.042352 *  
LOC_301                        6.839e-01  2.410e-01   2.838 0.004540 ** 
LOC_311                        1.360e-01  3.352e-01   0.406 0.685020    
LOC_321                        2.352e-01  2.446e-01   0.962 0.336282    
LOC_331                        2.013e-01  5.187e-01   0.388 0.697978    
LOC_341                       -1.789e-01  1.963e-01  -0.912 0.361961    
LOC_351                        4.760e-01  1.534e-01   3.103 0.001913 ** 
LOC_371                        7.133e-02  1.541e-01   0.463 0.643497    
LOC_381                        4.919e-01  3.669e-01   1.340 0.180084    
LOC_391                       -2.873e-01  1.444e-01  -1.989 0.046703 *  
LOC_401                        1.813e-01  1.672e-01   1.085 0.277980    
LOC_411                        1.185e-01  1.979e-01   0.599 0.549253    
LOC_421                       -2.917e-01  1.452e-01  -2.009 0.044519 *  
LOC_441                        3.419e-01  4.918e-01   0.695 0.486913    
LOC_461                        8.179e-01  2.548e-01   3.211 0.001324 ** 
LOC_471                       -1.756e-01  2.109e-01  -0.833 0.405121    
LOC_481                        2.550e-01  9.615e-02   2.652 0.008000 ** 
LOC_491                       -1.816e-01  2.097e-01  -0.866 0.386433    
LOC_501                       -4.056e-01  5.931e-01  -0.684 0.494059    
LOC_511                       -1.522e-01  8.253e-02  -1.844 0.065154 .  
LOC_531                        4.011e-01  1.306e-01   3.072 0.002126 ** 
LOC_551                       -3.177e-01  2.511e-01  -1.265 0.205694    
TOA_151                        6.730e-02  6.909e-02   0.974 0.329997    
TOA_201                        1.128e+00  1.928e-01   5.849 4.93e-09 ***
TOA_301                        3.468e-01  8.868e-02   3.910 9.22e-05 ***
TOA_321                        4.202e-01  7.540e-01   0.557 0.577355    
TOA_351                       -5.369e-01  2.865e-01  -1.874 0.060925 .  
TOA_381                        3.274e-01  6.717e-02   4.874 1.10e-06 ***
TOA_401                       -1.055e-01  1.763e-01  -0.598 0.549608    
TOA_421                        1.139e+00  4.414e-01   2.581 0.009837 ** 
TOA_441                        2.167e+00  1.072e+00   2.021 0.043296 *  
PPGROUP_111                   -4.104e-01  1.171e-01  -3.506 0.000455 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 20683  on 14919  degrees of freedom
Residual deviance: 14544  on 14851  degrees of freedom
AIC: 14682

Number of Fisher Scoring iterations: 5

                                   VIF
GSEGRD                        1.622738
SalaryOverUnderIndAvg         1.348899
LowerLimitAge                 2.319650
BLS_FEDERAL_OtherSep_Rate     1.779386
BLS_FEDERAL_Quits_Rate        1.238715
BLS_FEDERAL_JobOpenings_Level 1.188649
BLS_FEDERAL_Layoffs_Rate      1.400307
LOSSqrt                       1.856367
SEPCount_EFDATE_OCCLog        1.106791
AGELVL_B                      1.165822
AGELVL_C                      1.661625
AGELVL_E                      1.406918
AGELVL_F                      1.282980
AGELVL_G                      1.269558
AGELVL_H                      1.360004
AGELVL_I                      1.515946
LOC_01                        1.089328
LOC_02                        1.046139
LOC_04                        1.176431
LOC_05                        1.050457
LOC_06                        1.370739
LOC_08                        1.157708
LOC_10                        1.009095
LOC_12                        1.189254
LOC_13                        1.197700
LOC_15                        1.099290
LOC_17                        1.109069
LOC_18                        1.043379
LOC_19                        1.027585
LOC_20                        1.044017
LOC_22                        1.060106
LOC_24                        1.391427
LOC_25                        1.065889
LOC_26                        1.064928
LOC_27                        1.047198
LOC_28                        1.044135
LOC_29                        1.095148
LOC_30                        1.054410
LOC_31                        1.026741
LOC_32                        1.047077
LOC_33                        1.011737
LOC_34                        1.063960
LOC_35                        1.139519
LOC_37                        1.112600
LOC_38                        1.023370
LOC_39                        1.123051
LOC_40                        1.095179
LOC_41                        1.070114
LOC_42                        1.116521
LOC_44                        1.011471
LOC_46                        1.053036
LOC_47                        1.059537
LOC_48                        1.335241
LOC_49                        1.058759
LOC_50                        1.009143
LOC_51                        1.445043
LOC_53                        1.157121
LOC_55                        1.045923
TOA_15                        1.404428
TOA_20                        1.028182
TOA_30                        1.233026
TOA_32                        1.005917
TOA_35                        1.076566
TOA_38                        1.172315
TOA_40                        1.060716
TOA_42                        1.010484
TOA_44                        1.003670
PPGROUP_11                    1.155575

Removed AFTER this step: LOC_12 


Call:
glm(formula = fla, family = "binomial", data = OPMAnalysisDataNoFamBinary)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-2.8263  -0.7402  -0.1370   0.8187   3.2342  

Coefficients:
                                Estimate Std. Error z value Pr(>|z|)    
(Intercept)                    2.178e+00  3.397e-01   6.413 1.43e-10 ***
GSEGRD                        -6.365e-02  1.496e-02  -4.254 2.10e-05 ***
SalaryOverUnderIndAvg          7.841e-07  2.384e-06   0.329 0.742214    
LowerLimitAge                 -4.489e-02  2.654e-03 -16.915  < 2e-16 ***
BLS_FEDERAL_OtherSep_Rate      3.237e+00  3.374e-01   9.594  < 2e-16 ***
BLS_FEDERAL_Quits_Rate         4.745e-01  3.157e-01   1.503 0.132905    
BLS_FEDERAL_JobOpenings_Level  1.705e-02  1.929e-03   8.835  < 2e-16 ***
BLS_FEDERAL_Layoffs_Rate       4.351e-02  1.230e-01   0.354 0.723509    
LOSSqrt                       -7.822e-01  2.287e-02 -34.203  < 2e-16 ***
SEPCount_EFDATE_OCCLog        -3.136e-02  1.609e-02  -1.949 0.051304 .  
AGELVL_B1                     -5.842e-01  2.197e-01  -2.659 0.007836 ** 
AGELVL_C1                     -1.671e-01  8.532e-02  -1.959 0.050124 .  
AGELVL_E1                      2.741e-01  6.730e-02   4.072 4.65e-05 ***
AGELVL_F1                      4.900e-01  7.012e-02   6.988 2.78e-12 ***
AGELVL_G1                      7.134e-01  7.393e-02   9.649  < 2e-16 ***
AGELVL_H1                      8.479e-01  7.620e-02  11.127  < 2e-16 ***
AGELVL_I1                      7.188e-01  8.108e-02   8.865  < 2e-16 ***
LOC_011                       -2.132e-01  1.680e-01  -1.269 0.204416    
LOC_021                        1.515e-01  2.512e-01   0.603 0.546379    
LOC_041                        5.233e-01  1.380e-01   3.792 0.000150 ***
LOC_051                        7.797e-02  2.331e-01   0.334 0.738069    
LOC_061                        4.081e-01  8.966e-02   4.551 5.33e-06 ***
LOC_081                        1.408e-01  1.262e-01   1.116 0.264380    
LOC_101                        1.934e-01  5.811e-01   0.333 0.739279    
LOC_131                        1.661e-01  1.164e-01   1.428 0.153396    
LOC_151                       -7.155e-02  1.611e-01  -0.444 0.657011    
LOC_171                       -2.246e-01  1.447e-01  -1.553 0.120534    
LOC_181                       -2.310e-01  2.507e-01  -0.921 0.356844    
LOC_191                        2.116e-01  3.193e-01   0.663 0.507483    
LOC_201                        6.097e-01  2.558e-01   2.383 0.017151 *  
LOC_221                       -1.152e-01  2.075e-01  -0.555 0.578834    
LOC_241                       -1.120e-01  8.401e-02  -1.333 0.182469    
LOC_251                        1.410e-01  1.814e-01   0.777 0.437158    
LOC_261                       -7.502e-02  2.008e-01  -0.374 0.708646    
LOC_271                        2.848e-01  2.373e-01   1.200 0.230117    
LOC_281                       -1.289e-01  2.457e-01  -0.524 0.599934    
LOC_291                       -3.539e-01  1.692e-01  -2.092 0.036442 *  
LOC_301                        6.765e-01  2.398e-01   2.821 0.004787 ** 
LOC_311                        1.285e-01  3.343e-01   0.384 0.700804    
LOC_321                        2.279e-01  2.435e-01   0.936 0.349273    
LOC_331                        1.938e-01  5.181e-01   0.374 0.708349    
LOC_341                       -1.853e-01  1.952e-01  -0.949 0.342520    
LOC_351                        4.686e-01  1.516e-01   3.092 0.001989 ** 
LOC_371                        6.427e-02  1.525e-01   0.421 0.673391    
LOC_381                        4.845e-01  3.662e-01   1.323 0.185796    
LOC_391                       -2.943e-01  1.427e-01  -2.062 0.039225 *  
LOC_401                        1.743e-01  1.656e-01   1.052 0.292679    
LOC_411                        1.113e-01  1.966e-01   0.566 0.571235    
LOC_421                       -2.984e-01  1.436e-01  -2.078 0.037704 *  
LOC_441                        3.350e-01  4.912e-01   0.682 0.495282    
LOC_461                        8.104e-01  2.536e-01   3.195 0.001396 ** 
LOC_471                       -1.828e-01  2.096e-01  -0.872 0.383283    
LOC_481                        2.481e-01  9.359e-02   2.651 0.008035 ** 
LOC_491                       -1.888e-01  2.084e-01  -0.906 0.365092    
LOC_501                       -4.131e-01  5.926e-01  -0.697 0.485708    
LOC_511                       -1.587e-01  7.992e-02  -1.986 0.047086 *  
LOC_531                        3.943e-01  1.288e-01   3.062 0.002196 ** 
LOC_551                       -3.249e-01  2.500e-01  -1.300 0.193720    
TOA_151                        6.706e-02  6.908e-02   0.971 0.331681    
TOA_201                        1.127e+00  1.928e-01   5.848 4.98e-09 ***
TOA_301                        3.448e-01  8.846e-02   3.898 9.71e-05 ***
TOA_321                        4.204e-01  7.540e-01   0.558 0.577123    
TOA_351                       -5.383e-01  2.864e-01  -1.879 0.060201 .  
TOA_381                        3.283e-01  6.710e-02   4.892 9.99e-07 ***
TOA_401                       -1.068e-01  1.762e-01  -0.606 0.544492    
TOA_421                        1.138e+00  4.413e-01   2.579 0.009903 ** 
TOA_441                        2.161e+00  1.072e+00   2.016 0.043792 *  
PPGROUP_111                   -4.106e-01  1.171e-01  -3.508 0.000452 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 20683  on 14919  degrees of freedom
Residual deviance: 14544  on 14852  degrees of freedom
AIC: 14680

Number of Fisher Scoring iterations: 5

                                   VIF
GSEGRD                        1.607245
SalaryOverUnderIndAvg         1.335121
LowerLimitAge                 2.313769
BLS_FEDERAL_OtherSep_Rate     1.778650
BLS_FEDERAL_Quits_Rate        1.238726
BLS_FEDERAL_JobOpenings_Level 1.187870
BLS_FEDERAL_Layoffs_Rate      1.400016
LOSSqrt                       1.853161
SEPCount_EFDATE_OCCLog        1.106784
AGELVL_B                      1.165808
AGELVL_C                      1.661360
AGELVL_E                      1.406768
AGELVL_F                      1.282978
AGELVL_G                      1.268886
AGELVL_H                      1.359556
AGELVL_I                      1.515651
LOC_01                        1.070131
LOC_02                        1.038484
LOC_04                        1.142519
LOC_05                        1.039944
LOC_06                        1.303912
LOC_08                        1.124567
LOC_10                        1.007463
LOC_13                        1.155554
LOC_15                        1.080483
LOC_17                        1.086394
LOC_18                        1.034998
LOC_19                        1.021963
LOC_20                        1.035220
LOC_22                        1.047533
LOC_24                        1.319782
LOC_25                        1.052940
LOC_26                        1.052145
LOC_27                        1.037878
LOC_28                        1.034888
LOC_29                        1.075079
LOC_30                        1.044380
LOC_31                        1.021553
LOC_32                        1.037710
LOC_33                        1.009625
LOC_34                        1.052621
LOC_35                        1.112767
LOC_37                        1.089040
LOC_38                        1.019182
LOC_39                        1.096588
LOC_40                        1.075560
LOC_41                        1.055726
LOC_42                        1.092420
LOC_44                        1.009451
LOC_46                        1.043682
LOC_47                        1.047064
LOC_48                        1.265073
LOC_49                        1.046265
LOC_50                        1.007500
LOC_51                        1.355118
LOC_53                        1.125331
LOC_55                        1.037232
TOA_15                        1.404263
TOA_20                        1.028151
TOA_30                        1.226667
TOA_32                        1.005916
TOA_35                        1.076319
TOA_38                        1.170155
TOA_40                        1.060117
TOA_42                        1.010434
TOA_44                        1.003412
PPGROUP_11                    1.155529

Removed AFTER this step: SalaryOverUnderIndAvg 


Call:
glm(formula = fla, family = "binomial", data = OPMAnalysisDataNoFamBinary)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-2.8239  -0.7404  -0.1371   0.8187   3.2336  

Coefficients:
                               Estimate Std. Error z value Pr(>|z|)    
(Intercept)                    2.175776   0.339555   6.408 1.48e-10 ***
GSEGRD                        -0.064749   0.014587  -4.439 9.05e-06 ***
LowerLimitAge                 -0.044749   0.002619 -17.084  < 2e-16 ***
BLS_FEDERAL_OtherSep_Rate      3.237454   0.337409   9.595  < 2e-16 ***
BLS_FEDERAL_Quits_Rate         0.473729   0.315708   1.501 0.133478    
BLS_FEDERAL_JobOpenings_Level  0.017049   0.001929   8.836  < 2e-16 ***
BLS_FEDERAL_Layoffs_Rate       0.043437   0.122983   0.353 0.723942    
LOSSqrt                       -0.780266   0.022127 -35.263  < 2e-16 ***
SEPCount_EFDATE_OCCLog        -0.031296   0.016088  -1.945 0.051740 .  
AGELVL_B1                     -0.582401   0.219620  -2.652 0.008005 ** 
AGELVL_C1                     -0.166715   0.085303  -1.954 0.050655 .  
AGELVL_E1                      0.274455   0.067291   4.079 4.53e-05 ***
AGELVL_F1                      0.490885   0.070067   7.006 2.45e-12 ***
AGELVL_G1                      0.714289   0.073891   9.667  < 2e-16 ***
AGELVL_H1                      0.848498   0.076183  11.138  < 2e-16 ***
AGELVL_I1                      0.719356   0.081066   8.874  < 2e-16 ***
LOC_011                       -0.216571   0.167642  -1.292 0.196404    
LOC_021                        0.154810   0.251050   0.617 0.537466    
LOC_041                        0.518842   0.137333   3.778 0.000158 ***
LOC_051                        0.074755   0.232942   0.321 0.748273    
LOC_061                        0.412327   0.088709   4.648 3.35e-06 ***
LOC_081                        0.140018   0.126144   1.110 0.267008    
LOC_101                        0.191489   0.581466   0.329 0.741913    
LOC_131                        0.163159   0.116029   1.406 0.159665    
LOC_151                       -0.070196   0.161051  -0.436 0.662937    
LOC_171                       -0.224413   0.144687  -1.551 0.120896    
LOC_181                       -0.233289   0.250535  -0.931 0.351769    
LOC_191                        0.208357   0.319128   0.653 0.513823    
LOC_201                        0.605408   0.255417   2.370 0.017775 *  
LOC_221                       -0.117771   0.207374  -0.568 0.570091    
LOC_241                       -0.110760   0.083929  -1.320 0.186936    
LOC_251                        0.142889   0.181389   0.788 0.430843    
LOC_261                       -0.075152   0.200762  -0.374 0.708156    
LOC_271                        0.284589   0.237319   1.199 0.230457    
LOC_281                       -0.132894   0.245377  -0.542 0.588100    
LOC_291                       -0.357066   0.168919  -2.114 0.034530 *  
LOC_301                        0.675073   0.239642   2.817 0.004847 ** 
LOC_311                        0.124717   0.334130   0.373 0.708956    
LOC_321                        0.225988   0.243381   0.929 0.353131    
LOC_331                        0.192858   0.518035   0.372 0.709678    
LOC_341                       -0.182779   0.195126  -0.937 0.348902    
LOC_351                        0.464202   0.150974   3.075 0.002107 ** 
LOC_371                        0.062336   0.152365   0.409 0.682450    
LOC_381                        0.482358   0.366138   1.317 0.187698    
LOC_391                       -0.295863   0.142628  -2.074 0.038045 *  
LOC_401                        0.171745   0.165451   1.038 0.299249    
LOC_411                        0.109486   0.196519   0.557 0.577441    
LOC_421                       -0.299964   0.143524  -2.090 0.036619 *  
LOC_441                        0.335371   0.491192   0.683 0.494753    
LOC_461                        0.807237   0.253384   3.186 0.001443 ** 
LOC_471                       -0.185827   0.209412  -0.887 0.374877    
LOC_481                        0.247441   0.093566   2.645 0.008179 ** 
LOC_491                       -0.191339   0.208289  -0.919 0.358293    
LOC_501                       -0.418294   0.592240  -0.706 0.480007    
LOC_511                       -0.158525   0.079918  -1.984 0.047300 *  
LOC_531                        0.394822   0.128757   3.066 0.002167 ** 
LOC_551                       -0.326323   0.249927  -1.306 0.191664    
TOA_151                        0.066975   0.069074   0.970 0.332238    
TOA_201                        1.125992   0.192744   5.842 5.16e-09 ***
TOA_301                        0.346066   0.088379   3.916 9.01e-05 ***
TOA_321                        0.421249   0.754545   0.558 0.576652    
TOA_351                       -0.537868   0.286395  -1.878 0.060373 .  
TOA_381                        0.329542   0.066987   4.919 8.68e-07 ***
TOA_401                       -0.103990   0.176010  -0.591 0.554643    
TOA_421                        1.139177   0.441345   2.581 0.009847 ** 
TOA_441                        2.163222   1.072020   2.018 0.043602 *  
PPGROUP_111                   -0.407821   0.116710  -3.494 0.000475 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 20683  on 14919  degrees of freedom
Residual deviance: 14544  on 14853  degrees of freedom
AIC: 14678

Number of Fisher Scoring iterations: 5

                                   VIF
GSEGRD                        1.527142
LowerLimitAge                 2.254026
BLS_FEDERAL_OtherSep_Rate     1.778706
BLS_FEDERAL_Quits_Rate        1.238709
BLS_FEDERAL_JobOpenings_Level 1.187876
BLS_FEDERAL_Layoffs_Rate      1.400012
LOSSqrt                       1.735002
SEPCount_EFDATE_OCCLog        1.106562
AGELVL_B                      1.165066
AGELVL_C                      1.661044
AGELVL_E                      1.406387
AGELVL_F                      1.281202
AGELVL_G                      1.267262
AGELVL_H                      1.358699
AGELVL_I                      1.514945
LOC_01                        1.066097
LOC_02                        1.036863
LOC_04                        1.131292
LOC_05                        1.038118
LOC_06                        1.276608
LOC_08                        1.124136
LOC_10                        1.007351
LOC_13                        1.148525
LOC_15                        1.079801
LOC_17                        1.086368
LOC_18                        1.034179
LOC_19                        1.020988
LOC_20                        1.032491
LOC_22                        1.046036
LOC_24                        1.316978
LOC_25                        1.051856
LOC_26                        1.052143
LOC_27                        1.037870
LOC_28                        1.032325
LOC_29                        1.071707
LOC_30                        1.044110
LOC_31                        1.020370
LOC_32                        1.037130
LOC_33                        1.009601
LOC_34                        1.050963
LOC_35                        1.103842
LOC_37                        1.087433
LOC_38                        1.018865
LOC_39                        1.095332
LOC_40                        1.073209
LOC_41                        1.054876
LOC_42                        1.091195
LOC_44                        1.009445
LOC_46                        1.042212
LOC_47                        1.045038
LOC_48                        1.264614
LOC_49                        1.044829
LOC_50                        1.006794
LOC_51                        1.355129
LOC_53                        1.125190
LOC_55                        1.036940
TOA_15                        1.404316
TOA_20                        1.027572
TOA_30                        1.224091
TOA_32                        1.005899
TOA_35                        1.076276
TOA_38                        1.166152
TOA_40                        1.057631
TOA_42                        1.010401
TOA_44                        1.003387
PPGROUP_11                    1.149157

Removed AFTER this step: LOC_05 


Call:
glm(formula = fla, family = "binomial", data = OPMAnalysisDataNoFamBinary)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-2.8240  -0.7405  -0.1370   0.8188   3.2331  

Coefficients:
                               Estimate Std. Error z value Pr(>|z|)    
(Intercept)                    2.181947   0.339009   6.436 1.22e-10 ***
GSEGRD                        -0.064984   0.014568  -4.461 8.17e-06 ***
LowerLimitAge                 -0.044744   0.002619 -17.083  < 2e-16 ***
BLS_FEDERAL_OtherSep_Rate      3.235696   0.337361   9.591  < 2e-16 ***
BLS_FEDERAL_Quits_Rate         0.474361   0.315701   1.503 0.132951    
BLS_FEDERAL_JobOpenings_Level  0.017044   0.001929   8.834  < 2e-16 ***
BLS_FEDERAL_Layoffs_Rate       0.043687   0.122981   0.355 0.722413    
LOSSqrt                       -0.780194   0.022125 -35.263  < 2e-16 ***
SEPCount_EFDATE_OCCLog        -0.031332   0.016088  -1.948 0.051465 .  
AGELVL_B1                     -0.582689   0.219611  -2.653 0.007971 ** 
AGELVL_C1                     -0.166374   0.085297  -1.951 0.051113 .  
AGELVL_E1                      0.275001   0.067270   4.088 4.35e-05 ***
AGELVL_F1                      0.491499   0.070041   7.017 2.26e-12 ***
AGELVL_G1                      0.714360   0.073892   9.668  < 2e-16 ***
AGELVL_H1                      0.848880   0.076174  11.144  < 2e-16 ***
AGELVL_I1                      0.719882   0.081047   8.882  < 2e-16 ***
LOC_011                       -0.219559   0.167387  -1.312 0.189626    
LOC_021                        0.151665   0.250861   0.605 0.545460    
LOC_041                        0.515683   0.136979   3.765 0.000167 ***
LOC_061                        0.409304   0.088209   4.640 3.48e-06 ***
LOC_081                        0.137069   0.125812   1.089 0.275945    
LOC_101                        0.188118   0.581389   0.324 0.746267    
LOC_131                        0.160204   0.115664   1.385 0.166028    
LOC_151                       -0.073210   0.160777  -0.455 0.648857    
LOC_171                       -0.227338   0.144402  -1.574 0.115408    
LOC_181                       -0.236497   0.250339  -0.945 0.344808    
LOC_191                        0.205036   0.318969   0.643 0.520348    
LOC_201                        0.602291   0.255232   2.360 0.018286 *  
LOC_221                       -0.120900   0.207145  -0.584 0.559457    
LOC_241                       -0.113475   0.083503  -1.359 0.174169    
LOC_251                        0.139984   0.181163   0.773 0.439703    
LOC_261                       -0.078327   0.200520  -0.391 0.696079    
LOC_271                        0.281334   0.237109   1.187 0.235417    
LOC_281                       -0.136012   0.245184  -0.555 0.579076    
LOC_291                       -0.360232   0.168634  -2.136 0.032665 *  
LOC_301                        0.671848   0.239430   2.806 0.005016 ** 
LOC_311                        0.121456   0.333977   0.364 0.716107    
LOC_321                        0.222746   0.243177   0.916 0.359676    
LOC_331                        0.189409   0.517923   0.366 0.714582    
LOC_341                       -0.185757   0.194909  -0.953 0.340568    
LOC_351                        0.461130   0.150672   3.060 0.002210 ** 
LOC_371                        0.059257   0.152064   0.390 0.696769    
LOC_381                        0.479119   0.366004   1.309 0.190515    
LOC_391                       -0.298936   0.142308  -2.101 0.035674 *  
LOC_401                        0.168678   0.165173   1.021 0.307150    
LOC_411                        0.106294   0.196265   0.542 0.588106    
LOC_421                       -0.302888   0.143235  -2.115 0.034462 *  
LOC_441                        0.332335   0.491091   0.677 0.498578    
LOC_461                        0.803971   0.253183   3.175 0.001496 ** 
LOC_471                       -0.188954   0.209187  -0.903 0.366378    
LOC_481                        0.244406   0.093088   2.626 0.008651 ** 
LOC_491                       -0.194469   0.208063  -0.935 0.349963    
LOC_501                       -0.421394   0.592183  -0.712 0.476716    
LOC_511                       -0.161370   0.079428  -2.032 0.042188 *  
LOC_531                        0.391756   0.128402   3.051 0.002281 ** 
LOC_551                       -0.329610   0.249722  -1.320 0.186867    
TOA_151                        0.066807   0.069071   0.967 0.333433    
TOA_201                        1.125603   0.192739   5.840 5.22e-09 ***
TOA_301                        0.345155   0.088332   3.907 9.33e-05 ***
TOA_321                        0.421617   0.754508   0.559 0.576300    
TOA_351                       -0.539106   0.286368  -1.883 0.059760 .  
TOA_381                        0.330298   0.066947   4.934 8.07e-07 ***
TOA_401                       -0.104769   0.175998  -0.595 0.551652    
TOA_421                        1.138334   0.441339   2.579 0.009901 ** 
TOA_441                        2.160817   1.071983   2.016 0.043829 *  
PPGROUP_111                   -0.408162   0.116707  -3.497 0.000470 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 20683  on 14919  degrees of freedom
Residual deviance: 14544  on 14854  degrees of freedom
AIC: 14676

Number of Fisher Scoring iterations: 5

                                   VIF
GSEGRD                        1.523291
LowerLimitAge                 2.253889
BLS_FEDERAL_OtherSep_Rate     1.778212
BLS_FEDERAL_Quits_Rate        1.238671
BLS_FEDERAL_JobOpenings_Level 1.187809
BLS_FEDERAL_Layoffs_Rate      1.399937
LOSSqrt                       1.734824
SEPCount_EFDATE_OCCLog        1.106515
AGELVL_B                      1.165062
AGELVL_C                      1.660760
AGELVL_E                      1.405480
AGELVL_F                      1.280258
AGELVL_G                      1.267235
AGELVL_H                      1.358388
AGELVL_I                      1.514410
LOC_01                        1.062813
LOC_02                        1.035284
LOC_04                        1.125493
LOC_06                        1.262244
LOC_08                        1.118187
LOC_10                        1.007023
LOC_13                        1.141310
LOC_15                        1.076147
LOC_17                        1.082074
LOC_18                        1.032536
LOC_19                        1.019915
LOC_20                        1.031001
LOC_22                        1.043732
LOC_24                        1.303675
LOC_25                        1.049244
LOC_26                        1.049598
LOC_27                        1.035979
LOC_28                        1.030711
LOC_29                        1.068064
LOC_30                        1.042276
LOC_31                        1.019429
LOC_32                        1.035346
LOC_33                        1.009167
LOC_34                        1.048598
LOC_35                        1.099412
LOC_37                        1.083137
LOC_38                        1.018094
LOC_39                        1.090409
LOC_40                        1.069642
LOC_41                        1.052178
LOC_42                        1.086817
LOC_44                        1.009071
LOC_46                        1.040532
LOC_47                        1.042780
LOC_48                        1.251728
LOC_49                        1.042546
LOC_50                        1.006526
LOC_51                        1.338522
LOC_53                        1.119020
LOC_55                        1.035205
TOA_15                        1.404253
TOA_20                        1.027536
TOA_30                        1.222811
TOA_32                        1.005897
TOA_35                        1.076091
TOA_38                        1.164749
TOA_40                        1.057430
TOA_42                        1.010368
TOA_44                        1.003338
PPGROUP_11                    1.149065

Removed AFTER this step: LOC_10 


Call:
glm(formula = fla, family = "binomial", data = OPMAnalysisDataNoFamBinary)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-2.8240  -0.7404  -0.1370   0.8191   3.2328  

Coefficients:
                               Estimate Std. Error z value Pr(>|z|)    
(Intercept)                    2.183520   0.338986   6.441 1.18e-10 ***
GSEGRD                        -0.065072   0.014566  -4.467 7.92e-06 ***
LowerLimitAge                 -0.044732   0.002619 -17.080  < 2e-16 ***
BLS_FEDERAL_OtherSep_Rate      3.235607   0.337361   9.591  < 2e-16 ***
BLS_FEDERAL_Quits_Rate         0.474348   0.315695   1.503 0.132955    
BLS_FEDERAL_JobOpenings_Level  0.017046   0.001929   8.835  < 2e-16 ***
BLS_FEDERAL_Layoffs_Rate       0.043330   0.122976   0.352 0.724576    
LOSSqrt                       -0.780185   0.022125 -35.262  < 2e-16 ***
SEPCount_EFDATE_OCCLog        -0.031363   0.016088  -1.949 0.051240 .  
AGELVL_B1                     -0.583101   0.219612  -2.655 0.007927 ** 
AGELVL_C1                     -0.166364   0.085298  -1.950 0.051129 .  
AGELVL_E1                      0.275066   0.067270   4.089 4.33e-05 ***
AGELVL_F1                      0.491845   0.070032   7.023 2.17e-12 ***
AGELVL_G1                      0.714311   0.073892   9.667  < 2e-16 ***
AGELVL_H1                      0.848861   0.076173  11.144  < 2e-16 ***
AGELVL_I1                      0.720007   0.081043   8.884  < 2e-16 ***
LOC_011                       -0.220715   0.167351  -1.319 0.187212    
LOC_021                        0.150362   0.250835   0.599 0.548876    
LOC_041                        0.514379   0.136920   3.757 0.000172 ***
LOC_061                        0.408107   0.088132   4.631 3.65e-06 ***
LOC_081                        0.135873   0.125759   1.080 0.279954    
LOC_131                        0.159039   0.115609   1.376 0.168925    
LOC_151                       -0.074416   0.160735  -0.463 0.643383    
LOC_171                       -0.228463   0.144361  -1.583 0.113517    
LOC_181                       -0.237840   0.250308  -0.950 0.342016    
LOC_191                        0.203711   0.318947   0.639 0.523018    
LOC_201                        0.600993   0.255204   2.355 0.018525 *  
LOC_221                       -0.122169   0.207111  -0.590 0.555276    
LOC_241                       -0.114548   0.083438  -1.373 0.169798    
LOC_251                        0.138850   0.181130   0.767 0.443333    
LOC_261                       -0.079596   0.200483  -0.397 0.691351    
LOC_271                        0.280026   0.237079   1.181 0.237543    
LOC_281                       -0.137292   0.245156  -0.560 0.575467    
LOC_291                       -0.361475   0.168593  -2.144 0.032027 *  
LOC_301                        0.670537   0.239398   2.801 0.005096 ** 
LOC_311                        0.120175   0.333956   0.360 0.718957    
LOC_321                        0.221361   0.243143   0.910 0.362604    
LOC_331                        0.188021   0.517907   0.363 0.716575    
LOC_341                       -0.186883   0.194879  -0.959 0.337576    
LOC_351                        0.459903   0.150625   3.053 0.002263 ** 
LOC_371                        0.058013   0.152017   0.382 0.702741    
LOC_381                        0.477802   0.365988   1.306 0.191718    
LOC_391                       -0.300136   0.142262  -2.110 0.034880 *  
LOC_401                        0.167461   0.165131   1.014 0.310531    
LOC_411                        0.105014   0.196225   0.535 0.592532    
LOC_421                       -0.304008   0.143195  -2.123 0.033751 *  
LOC_441                        0.331101   0.491071   0.674 0.500158    
LOC_461                        0.802644   0.253153   3.171 0.001521 ** 
LOC_471                       -0.190244   0.209152  -0.910 0.363034    
LOC_481                        0.243190   0.093013   2.615 0.008933 ** 
LOC_491                       -0.195732   0.208031  -0.941 0.346767    
LOC_501                       -0.422795   0.592174  -0.714 0.475245    
LOC_511                       -0.162508   0.079352  -2.048 0.040566 *  
LOC_531                        0.390564   0.128349   3.043 0.002342 ** 
LOC_551                       -0.330917   0.249692  -1.325 0.185072    
TOA_151                        0.067272   0.069058   0.974 0.329988    
TOA_201                        1.125437   0.192738   5.839 5.24e-09 ***
TOA_301                        0.344844   0.088326   3.904 9.45e-05 ***
TOA_321                        0.421811   0.754476   0.559 0.576109    
TOA_351                       -0.539355   0.286369  -1.883 0.059643 .  
TOA_381                        0.330795   0.066927   4.943 7.71e-07 ***
TOA_401                       -0.104925   0.175998  -0.596 0.551059    
TOA_421                        1.138093   0.441338   2.579 0.009916 ** 
TOA_441                        2.159903   1.071975   2.015 0.043917 *  
PPGROUP_111                   -0.407968   0.116706  -3.496 0.000473 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 20683  on 14919  degrees of freedom
Residual deviance: 14544  on 14855  degrees of freedom
AIC: 14674

Number of Fisher Scoring iterations: 5

                                   VIF
GSEGRD                        1.522785
LowerLimitAge                 2.253353
BLS_FEDERAL_OtherSep_Rate     1.778210
BLS_FEDERAL_Quits_Rate        1.238664
BLS_FEDERAL_JobOpenings_Level 1.187809
BLS_FEDERAL_Layoffs_Rate      1.399811
LOSSqrt                       1.734807
SEPCount_EFDATE_OCCLog        1.106482
AGELVL_B                      1.165040
AGELVL_C                      1.660770
AGELVL_E                      1.405477
AGELVL_F                      1.280008
AGELVL_G                      1.267222
AGELVL_H                      1.358360
AGELVL_I                      1.514380
LOC_01                        1.062331
LOC_02                        1.035017
LOC_04                        1.124519
LOC_06                        1.260043
LOC_08                        1.117227
LOC_13                        1.140212
LOC_15                        1.075572
LOC_17                        1.081454
LOC_18                        1.032254
LOC_19                        1.019748
LOC_20                        1.030746
LOC_22                        1.043362
LOC_24                        1.301641
LOC_25                        1.048856
LOC_26                        1.049201
LOC_27                        1.035679
LOC_28                        1.030445
LOC_29                        1.067514
LOC_30                        1.041979
LOC_31                        1.019288
LOC_32                        1.035027
LOC_33                        1.009098
LOC_34                        1.048271
LOC_35                        1.098716
LOC_37                        1.082450
LOC_38                        1.017969
LOC_39                        1.089676
LOC_40                        1.069092
LOC_41                        1.051757
LOC_42                        1.086193
LOC_44                        1.009011
LOC_46                        1.040257
LOC_47                        1.042404
LOC_48                        1.249699
LOC_49                        1.042181
LOC_50                        1.006473
LOC_51                        1.335913
LOC_53                        1.118110
LOC_55                        1.034938
TOA_15                        1.403746
TOA_20                        1.027529
TOA_30                        1.222663
TOA_32                        1.005897
TOA_35                        1.076098
TOA_38                        1.164156
TOA_40                        1.057424
TOA_42                        1.010366
TOA_44                        1.003331
PPGROUP_11                    1.149045

Removed AFTER this step: BLS_FEDERAL_Layoffs_Rate 


Call:
glm(formula = fla, family = "binomial", data = OPMAnalysisDataNoFamBinary)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-2.8215  -0.7397  -0.1370   0.8186   3.2323  

Coefficients:
                               Estimate Std. Error z value Pr(>|z|)    
(Intercept)                    2.168303   0.336227   6.449 1.13e-10 ***
GSEGRD                        -0.065014   0.014565  -4.464 8.05e-06 ***
LowerLimitAge                 -0.044721   0.002619 -17.079  < 2e-16 ***
BLS_FEDERAL_OtherSep_Rate      3.294584   0.292911  11.248  < 2e-16 ***
BLS_FEDERAL_Quits_Rate         0.458451   0.312495   1.467 0.142358    
BLS_FEDERAL_JobOpenings_Level  0.017241   0.001848   9.329  < 2e-16 ***
LOSSqrt                       -0.779860   0.022105 -35.280  < 2e-16 ***
SEPCount_EFDATE_OCCLog        -0.031162   0.016078  -1.938 0.052599 .  
AGELVL_B1                     -0.583027   0.219636  -2.655 0.007942 ** 
AGELVL_C1                     -0.166363   0.085301  -1.950 0.051139 .  
AGELVL_E1                      0.275089   0.067271   4.089 4.33e-05 ***
AGELVL_F1                      0.491414   0.070019   7.018 2.25e-12 ***
AGELVL_G1                      0.713908   0.073879   9.663  < 2e-16 ***
AGELVL_H1                      0.848290   0.076152  11.139  < 2e-16 ***
AGELVL_I1                      0.719584   0.081035   8.880  < 2e-16 ***
LOC_011                       -0.221438   0.167386  -1.323 0.185864    
LOC_021                        0.149818   0.250785   0.597 0.550242    
LOC_041                        0.513942   0.136919   3.754 0.000174 ***
LOC_061                        0.408131   0.088133   4.631 3.64e-06 ***
LOC_081                        0.135394   0.125749   1.077 0.281616    
LOC_131                        0.158738   0.115600   1.373 0.169700    
LOC_151                       -0.074376   0.160743  -0.463 0.643577    
LOC_171                       -0.228818   0.144361  -1.585 0.112958    
LOC_181                       -0.238110   0.250331  -0.951 0.341512    
LOC_191                        0.204350   0.319001   0.641 0.521787    
LOC_201                        0.602332   0.255176   2.360 0.018252 *  
LOC_221                       -0.121545   0.207086  -0.587 0.557252    
LOC_241                       -0.115021   0.083428  -1.379 0.167991    
LOC_251                        0.139073   0.181104   0.768 0.442535    
LOC_261                       -0.079572   0.200498  -0.397 0.691462    
LOC_271                        0.279427   0.237010   1.179 0.238410    
LOC_281                       -0.136989   0.245126  -0.559 0.576264    
LOC_291                       -0.360644   0.168620  -2.139 0.032452 *  
LOC_301                        0.670778   0.239380   2.802 0.005076 ** 
LOC_311                        0.120621   0.333958   0.361 0.717960    
LOC_321                        0.220598   0.243126   0.907 0.364226    
LOC_331                        0.187430   0.517922   0.362 0.717435    
LOC_341                       -0.185904   0.194873  -0.954 0.340096    
LOC_351                        0.459787   0.150637   3.052 0.002271 ** 
LOC_371                        0.057762   0.152012   0.380 0.703957    
LOC_381                        0.477269   0.366038   1.304 0.192275    
LOC_391                       -0.299478   0.142245  -2.105 0.035259 *  
LOC_401                        0.167606   0.165098   1.015 0.310015    
LOC_411                        0.105391   0.196253   0.537 0.591258    
LOC_421                       -0.304551   0.143194  -2.127 0.033433 *  
LOC_441                        0.331390   0.491077   0.675 0.499788    
LOC_461                        0.802590   0.253180   3.170 0.001524 ** 
LOC_471                       -0.191428   0.209120  -0.915 0.359983    
LOC_481                        0.243378   0.093008   2.617 0.008877 ** 
LOC_491                       -0.195817   0.208006  -0.941 0.346499    
LOC_501                       -0.425532   0.591928  -0.719 0.472208    
LOC_511                       -0.162115   0.079342  -2.043 0.041028 *  
LOC_531                        0.390516   0.128340   3.043 0.002344 ** 
LOC_551                       -0.330474   0.249731  -1.323 0.185730    
TOA_151                        0.067497   0.069057   0.977 0.328362    
TOA_201                        1.125809   0.192719   5.842 5.17e-09 ***
TOA_301                        0.344702   0.088320   3.903 9.51e-05 ***
TOA_321                        0.417433   0.754276   0.553 0.579975    
TOA_351                       -0.538782   0.286395  -1.881 0.059937 .  
TOA_381                        0.330875   0.066926   4.944 7.66e-07 ***
TOA_401                       -0.105467   0.175986  -0.599 0.548978    
TOA_421                        1.139273   0.441399   2.581 0.009850 ** 
TOA_441                        2.159821   1.071976   2.015 0.043925 *  
PPGROUP_111                   -0.407455   0.116689  -3.492 0.000480 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 20683  on 14919  degrees of freedom
Residual deviance: 14544  on 14856  degrees of freedom
AIC: 14672

Number of Fisher Scoring iterations: 5

                                   VIF
GSEGRD                        1.522526
LowerLimitAge                 2.252678
BLS_FEDERAL_OtherSep_Rate     1.341059
BLS_FEDERAL_Quits_Rate        1.213219
BLS_FEDERAL_JobOpenings_Level 1.090067
LOSSqrt                       1.731243
SEPCount_EFDATE_OCCLog        1.105098
AGELVL_B                      1.164971
AGELVL_C                      1.660784
AGELVL_E                      1.405500
AGELVL_F                      1.279617
AGELVL_G                      1.266914
AGELVL_H                      1.357709
AGELVL_I                      1.514000
LOC_01                        1.062121
LOC_02                        1.034990
LOC_04                        1.124417
LOC_06                        1.260015
LOC_08                        1.117082
LOC_13                        1.140145
LOC_15                        1.075570
LOC_17                        1.081393
LOC_18                        1.032244
LOC_19                        1.019707
LOC_20                        1.030516
LOC_22                        1.043295
LOC_24                        1.301291
LOC_25                        1.048858
LOC_26                        1.049178
LOC_27                        1.035641
LOC_28                        1.030433
LOC_29                        1.067263
LOC_30                        1.041969
LOC_31                        1.019269
LOC_32                        1.034950
LOC_33                        1.009094
LOC_34                        1.048058
LOC_35                        1.098709
LOC_37                        1.082421
LOC_38                        1.017943
LOC_39                        1.089478
LOC_40                        1.069107
LOC_41                        1.051718
LOC_42                        1.086052
LOC_44                        1.009007
LOC_46                        1.040244
LOC_47                        1.042137
LOC_48                        1.249661
LOC_49                        1.042196
LOC_50                        1.006303
LOC_51                        1.335636
LOC_53                        1.118108
LOC_55                        1.034902
TOA_15                        1.403673
TOA_20                        1.027479
TOA_30                        1.222618
TOA_32                        1.005622
TOA_35                        1.076003
TOA_38                        1.164145
TOA_40                        1.057360
TOA_42                        1.010300
TOA_44                        1.003330
PPGROUP_11                    1.148870

Removed AFTER this step: LOC_31 


Call:
glm(formula = fla, family = "binomial", data = OPMAnalysisDataNoFamBinary)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-2.8220  -0.7400  -0.1370   0.8187   3.2319  

Coefficients:
                               Estimate Std. Error z value Pr(>|z|)    
(Intercept)                    2.174834   0.335751   6.478 9.32e-11 ***
GSEGRD                        -0.065270   0.014548  -4.487 7.24e-06 ***
LowerLimitAge                 -0.044707   0.002618 -17.075  < 2e-16 ***
BLS_FEDERAL_OtherSep_Rate      3.293663   0.292898  11.245  < 2e-16 ***
BLS_FEDERAL_Quits_Rate         0.459655   0.312479   1.471 0.141292    
BLS_FEDERAL_JobOpenings_Level  0.017229   0.001848   9.324  < 2e-16 ***
LOSSqrt                       -0.779830   0.022105 -35.279  < 2e-16 ***
SEPCount_EFDATE_OCCLog        -0.031322   0.016072  -1.949 0.051309 .  
AGELVL_B1                     -0.583756   0.219630  -2.658 0.007863 ** 
AGELVL_C1                     -0.166259   0.085298  -1.949 0.051278 .  
AGELVL_E1                      0.275588   0.067257   4.098 4.18e-05 ***
AGELVL_F1                      0.491564   0.070017   7.021 2.21e-12 ***
AGELVL_G1                      0.714140   0.073875   9.667  < 2e-16 ***
AGELVL_H1                      0.848168   0.076151  11.138  < 2e-16 ***
AGELVL_I1                      0.719631   0.081037   8.880  < 2e-16 ***
LOC_011                       -0.223687   0.167273  -1.337 0.181139    
LOC_021                        0.147296   0.250692   0.588 0.556828    
LOC_041                        0.511489   0.136750   3.740 0.000184 ***
LOC_061                        0.405804   0.087898   4.617 3.90e-06 ***
LOC_081                        0.133079   0.125587   1.060 0.289301    
LOC_131                        0.156470   0.115430   1.356 0.175245    
LOC_151                       -0.076728   0.160612  -0.478 0.632847    
LOC_171                       -0.231045   0.144230  -1.602 0.109173    
LOC_181                       -0.240410   0.250248  -0.961 0.336709    
LOC_191                        0.201875   0.318920   0.633 0.526737    
LOC_201                        0.599908   0.255088   2.352 0.018684 *  
LOC_221                       -0.124012   0.206975  -0.599 0.549063    
LOC_241                       -0.117110   0.083228  -1.407 0.159397    
LOC_251                        0.136890   0.181002   0.756 0.449477    
LOC_261                       -0.081922   0.200394  -0.409 0.682684    
LOC_271                        0.276968   0.236914   1.169 0.242377    
LOC_281                       -0.139363   0.245034  -0.569 0.569528    
LOC_291                       -0.363032   0.168492  -2.155 0.031193 *  
LOC_301                        0.668152   0.239267   2.792 0.005230 ** 
LOC_321                        0.218141   0.243027   0.898 0.369400    
LOC_331                        0.184898   0.517887   0.357 0.721074    
LOC_341                       -0.188135   0.194778  -0.966 0.334096    
LOC_351                        0.457378   0.150490   3.039 0.002372 ** 
LOC_371                        0.055434   0.151875   0.365 0.715111    
LOC_381                        0.474504   0.365955   1.297 0.194762    
LOC_391                       -0.301770   0.142105  -2.124 0.033706 *  
LOC_401                        0.165250   0.164968   1.002 0.316483    
LOC_411                        0.102862   0.196128   0.524 0.599954    
LOC_421                       -0.306760   0.143064  -2.144 0.032016 *  
LOC_441                        0.328955   0.491012   0.670 0.502888    
LOC_461                        0.800001   0.253077   3.161 0.001572 ** 
LOC_471                       -0.193804   0.209018  -0.927 0.353817    
LOC_481                        0.241054   0.092786   2.598 0.009378 ** 
LOC_491                       -0.198217   0.207902  -0.953 0.340378    
LOC_501                       -0.427811   0.591872  -0.723 0.469796    
LOC_511                       -0.164311   0.079112  -2.077 0.037805 *  
LOC_531                        0.388116   0.128168   3.028 0.002460 ** 
LOC_551                       -0.332900   0.249639  -1.334 0.182360    
TOA_151                        0.067301   0.069053   0.975 0.329748    
TOA_201                        1.126230   0.192721   5.844 5.10e-09 ***
TOA_301                        0.344006   0.088299   3.896 9.78e-05 ***
TOA_321                        0.417389   0.754249   0.553 0.580001    
TOA_351                       -0.539989   0.286379  -1.886 0.059352 .  
TOA_381                        0.330707   0.066921   4.942 7.74e-07 ***
TOA_401                       -0.106096   0.175983  -0.603 0.546591    
TOA_421                        1.141237   0.441298   2.586 0.009707 ** 
TOA_441                        2.158296   1.071961   2.013 0.044072 *  
PPGROUP_111                   -0.408003   0.116674  -3.497 0.000471 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 20683  on 14919  degrees of freedom
Residual deviance: 14544  on 14857  degrees of freedom
AIC: 14670

Number of Fisher Scoring iterations: 5

                                   VIF
GSEGRD                        1.518890
LowerLimitAge                 2.252156
BLS_FEDERAL_OtherSep_Rate     1.340928
BLS_FEDERAL_Quits_Rate        1.213060
BLS_FEDERAL_JobOpenings_Level 1.089716
LOSSqrt                       1.731182
SEPCount_EFDATE_OCCLog        1.104275
AGELVL_B                      1.164884
AGELVL_C                      1.660730
AGELVL_E                      1.404908
AGELVL_F                      1.279585
AGELVL_G                      1.266846
AGELVL_H                      1.357674
AGELVL_I                      1.513987
LOC_01                        1.060657
LOC_02                        1.034190
LOC_04                        1.121670
LOC_06                        1.253316
LOC_08                        1.114198
LOC_13                        1.136800
LOC_15                        1.073821
LOC_17                        1.079438
LOC_18                        1.031580
LOC_19                        1.019240
LOC_20                        1.029806
LOC_22                        1.042166
LOC_24                        1.295076
LOC_25                        1.047698
LOC_26                        1.048083
LOC_27                        1.034792
LOC_28                        1.029699
LOC_29                        1.065633
LOC_30                        1.041013
LOC_32                        1.034146
LOC_33                        1.008909
LOC_34                        1.047016
LOC_35                        1.096568
LOC_37                        1.080494
LOC_38                        1.017498
LOC_39                        1.087323
LOC_40                        1.067453
LOC_41                        1.050386
LOC_42                        1.084089
LOC_44                        1.008818
LOC_46                        1.039417
LOC_47                        1.041114
LOC_48                        1.243719
LOC_49                        1.041140
LOC_50                        1.006190
LOC_51                        1.327837
LOC_53                        1.115132
LOC_55                        1.034160
TOA_15                        1.403614
TOA_20                        1.027442
TOA_30                        1.222028
TOA_32                        1.005622
TOA_35                        1.075881
TOA_38                        1.164063
TOA_40                        1.057257
TOA_42                        1.010168
TOA_44                        1.003315
PPGROUP_11                    1.148685

Removed AFTER this step: LOC_33 


Call:
glm(formula = fla, family = "binomial", data = OPMAnalysisDataNoFamBinary)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-2.8220  -0.7398  -0.1370   0.8188   3.2316  

Coefficients:
                               Estimate Std. Error z value Pr(>|z|)    
(Intercept)                    2.178149   0.335638   6.490 8.61e-11 ***
GSEGRD                        -0.065393   0.014544  -4.496 6.92e-06 ***
LowerLimitAge                 -0.044698   0.002618 -17.074  < 2e-16 ***
BLS_FEDERAL_OtherSep_Rate      3.292928   0.292892  11.243  < 2e-16 ***
BLS_FEDERAL_Quits_Rate         0.459667   0.312468   1.471 0.141268    
BLS_FEDERAL_JobOpenings_Level  0.017219   0.001848   9.319  < 2e-16 ***
LOSSqrt                       -0.779791   0.022104 -35.278  < 2e-16 ***
SEPCount_EFDATE_OCCLog        -0.031339   0.016072  -1.950 0.051188 .  
AGELVL_B1                     -0.582697   0.219610  -2.653 0.007970 ** 
AGELVL_C1                     -0.166112   0.085298  -1.947 0.051483 .  
AGELVL_E1                      0.275630   0.067256   4.098 4.16e-05 ***
AGELVL_F1                      0.491601   0.070015   7.021 2.20e-12 ***
AGELVL_G1                      0.714097   0.073874   9.666  < 2e-16 ***
AGELVL_H1                      0.848157   0.076151  11.138  < 2e-16 ***
AGELVL_I1                      0.719640   0.081034   8.881  < 2e-16 ***
LOC_011                       -0.225152   0.167226  -1.346 0.178176    
LOC_021                        0.145853   0.250657   0.582 0.560647    
LOC_041                        0.510020   0.136686   3.731 0.000190 ***
LOC_061                        0.404343   0.087804   4.605 4.12e-06 ***
LOC_081                        0.131719   0.125530   1.049 0.294036    
LOC_131                        0.155114   0.115368   1.345 0.178780    
LOC_151                       -0.078094   0.160565  -0.486 0.626705    
LOC_171                       -0.232442   0.144178  -1.612 0.106922    
LOC_181                       -0.242022   0.250208  -0.967 0.333402    
LOC_191                        0.200212   0.318896   0.628 0.530116    
LOC_201                        0.598435   0.255057   2.346 0.018962 *  
LOC_221                       -0.125495   0.206933  -0.606 0.544214    
LOC_241                       -0.118327   0.083159  -1.423 0.154763    
LOC_251                        0.135506   0.180962   0.749 0.453974    
LOC_261                       -0.083512   0.200345  -0.417 0.676794    
LOC_271                        0.275370   0.236878   1.163 0.245032    
LOC_281                       -0.140861   0.244998  -0.575 0.565327    
LOC_291                       -0.364553   0.168441  -2.164 0.030443 *  
LOC_301                        0.666624   0.239227   2.787 0.005327 ** 
LOC_321                        0.216495   0.242986   0.891 0.372941    
LOC_341                       -0.189570   0.194739  -0.973 0.330325    
LOC_351                        0.455938   0.150436   3.031 0.002439 ** 
LOC_371                        0.053964   0.151819   0.355 0.722253    
LOC_381                        0.472926   0.365933   1.292 0.196225    
LOC_391                       -0.303293   0.142043  -2.135 0.032743 *  
LOC_401                        0.163858   0.164920   0.994 0.320436    
LOC_411                        0.101301   0.196078   0.517 0.605409    
LOC_421                       -0.308226   0.143007  -2.155 0.031137 *  
LOC_441                        0.327418   0.490988   0.667 0.504864    
LOC_461                        0.798439   0.253041   3.155 0.001603 ** 
LOC_471                       -0.195356   0.208974  -0.935 0.349873    
LOC_481                        0.239630   0.092700   2.585 0.009738 ** 
LOC_491                       -0.199677   0.207861  -0.961 0.336738    
LOC_501                       -0.429401   0.591866  -0.726 0.468144    
LOC_511                       -0.165605   0.079030  -2.095 0.036129 *  
LOC_531                        0.386676   0.128104   3.018 0.002541 ** 
LOC_551                       -0.334612   0.249600  -1.341 0.180053    
TOA_151                        0.067115   0.069050   0.972 0.331057    
TOA_201                        1.125945   0.192719   5.842 5.14e-09 ***
TOA_301                        0.343578   0.088290   3.891 9.96e-05 ***
TOA_321                        0.417539   0.754225   0.554 0.579852    
TOA_351                       -0.540798   0.286368  -1.888 0.058963 .  
TOA_381                        0.331547   0.066882   4.957 7.15e-07 ***
TOA_401                       -0.106421   0.175983  -0.605 0.545364    
TOA_421                        1.143755   0.441206   2.592 0.009533 ** 
TOA_441                        2.157351   1.071955   2.013 0.044163 *  
PPGROUP_111                   -0.407918   0.116673  -3.496 0.000472 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 20683  on 14919  degrees of freedom
Residual deviance: 14545  on 14858  degrees of freedom
AIC: 14669

Number of Fisher Scoring iterations: 5

                                   VIF
GSEGRD                        1.518055
LowerLimitAge                 2.251755
BLS_FEDERAL_OtherSep_Rate     1.340828
BLS_FEDERAL_Quits_Rate        1.213043
BLS_FEDERAL_JobOpenings_Level 1.089487
LOSSqrt                       1.731028
SEPCount_EFDATE_OCCLog        1.104286
AGELVL_B                      1.164643
AGELVL_C                      1.660683
AGELVL_E                      1.404911
AGELVL_F                      1.279578
AGELVL_G                      1.266801
AGELVL_H                      1.357583
AGELVL_I                      1.513874
LOC_01                        1.060022
LOC_02                        1.033925
LOC_04                        1.120672
LOC_06                        1.250628
LOC_08                        1.113189
LOC_13                        1.135586
LOC_15                        1.073225
LOC_17                        1.078652
LOC_18                        1.031249
LOC_19                        1.019023
LOC_20                        1.029539
LOC_22                        1.041753
LOC_24                        1.292937
LOC_25                        1.047222
LOC_26                        1.047574
LOC_27                        1.034425
LOC_28                        1.029400
LOC_29                        1.064960
LOC_30                        1.040684
LOC_32                        1.033776
LOC_34                        1.046575
LOC_35                        1.095791
LOC_37                        1.079713
LOC_38                        1.017348
LOC_39                        1.086352
LOC_40                        1.066869
LOC_41                        1.049868
LOC_42                        1.083202
LOC_44                        1.008740
LOC_46                        1.039109
LOC_47                        1.040667
LOC_48                        1.241450
LOC_49                        1.040743
LOC_50                        1.006133
LOC_51                        1.325094
LOC_53                        1.114044
LOC_55                        1.033783
TOA_15                        1.403502
TOA_20                        1.027424
TOA_30                        1.221805
TOA_32                        1.005622
TOA_35                        1.075805
TOA_38                        1.162650
TOA_40                        1.057228
TOA_42                        1.009962
TOA_44                        1.003309
PPGROUP_11                    1.148697

Removed AFTER this step: LOC_37 


Call:
glm(formula = fla, family = "binomial", data = OPMAnalysisDataNoFamBinary)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-2.8219  -0.7402  -0.1370   0.8190   3.2308  

Coefficients:
                               Estimate Std. Error z value Pr(>|z|)    
(Intercept)                    2.184472   0.335177   6.517 7.15e-11 ***
GSEGRD                        -0.065714   0.014518  -4.526 6.00e-06 ***
LowerLimitAge                 -0.044682   0.002618 -17.070  < 2e-16 ***
BLS_FEDERAL_OtherSep_Rate      3.292797   0.292891  11.242  < 2e-16 ***
BLS_FEDERAL_Quits_Rate         0.463031   0.312323   1.483 0.138197    
BLS_FEDERAL_JobOpenings_Level  0.017214   0.001848   9.317  < 2e-16 ***
LOSSqrt                       -0.779881   0.022103 -35.284  < 2e-16 ***
SEPCount_EFDATE_OCCLog        -0.031132   0.016061  -1.938 0.052585 .  
AGELVL_B1                     -0.583580   0.219585  -2.658 0.007869 ** 
AGELVL_C1                     -0.166522   0.085290  -1.952 0.050888 .  
AGELVL_E1                      0.275812   0.067254   4.101 4.11e-05 ***
AGELVL_F1                      0.491695   0.070012   7.023 2.17e-12 ***
AGELVL_G1                      0.714074   0.073873   9.666  < 2e-16 ***
AGELVL_H1                      0.848405   0.076148  11.142  < 2e-16 ***
AGELVL_I1                      0.719540   0.081037   8.879  < 2e-16 ***
LOC_011                       -0.229783   0.166725  -1.378 0.168137    
LOC_021                        0.140821   0.250270   0.563 0.573657    
LOC_041                        0.504960   0.135944   3.714 0.000204 ***
LOC_061                        0.399635   0.086803   4.604 4.15e-06 ***
LOC_081                        0.126955   0.124819   1.017 0.309097    
LOC_131                        0.150404   0.114609   1.312 0.189409    
LOC_151                       -0.082994   0.159976  -0.519 0.603908    
LOC_171                       -0.237067   0.143594  -1.651 0.098747 .  
LOC_181                       -0.246821   0.249846  -0.988 0.323204    
LOC_191                        0.195402   0.318614   0.613 0.539686    
LOC_201                        0.593483   0.254692   2.330 0.019796 *  
LOC_221                       -0.130318   0.206493  -0.631 0.527974    
LOC_241                       -0.122670   0.082260  -1.491 0.135896    
LOC_251                        0.131018   0.180527   0.726 0.467989    
LOC_261                       -0.088415   0.199881  -0.442 0.658247    
LOC_271                        0.270432   0.236480   1.144 0.252801    
LOC_281                       -0.145666   0.244637  -0.595 0.551549    
LOC_291                       -0.369413   0.167892  -2.200 0.027785 *  
LOC_301                        0.661550   0.238811   2.770 0.005602 ** 
LOC_321                        0.211692   0.242619   0.873 0.382920    
LOC_341                       -0.194109   0.194326  -0.999 0.317853    
LOC_351                        0.450928   0.149777   3.011 0.002607 ** 
LOC_381                        0.467967   0.365692   1.280 0.200659    
LOC_391                       -0.307993   0.141431  -2.178 0.029430 *  
LOC_401                        0.159026   0.164359   0.968 0.333270    
LOC_411                        0.096550   0.195625   0.494 0.621625    
LOC_421                       -0.312803   0.142428  -2.196 0.028076 *  
LOC_441                        0.322616   0.490774   0.657 0.510948    
LOC_461                        0.793074   0.252596   3.140 0.001691 ** 
LOC_471                       -0.200193   0.208536  -0.960 0.337058    
LOC_481                        0.234751   0.091682   2.560 0.010453 *  
LOC_491                       -0.204432   0.207434  -0.986 0.324364    
LOC_501                       -0.434257   0.591725  -0.734 0.463020    
LOC_511                       -0.170166   0.077988  -2.182 0.029112 *  
LOC_531                        0.381892   0.127398   2.998 0.002721 ** 
LOC_551                       -0.339445   0.249239  -1.362 0.173222    
TOA_151                        0.066885   0.069045   0.969 0.332684    
TOA_201                        1.125044   0.192710   5.838 5.28e-09 ***
TOA_301                        0.341710   0.088139   3.877 0.000106 ***
TOA_321                        0.417849   0.754189   0.554 0.579553    
TOA_351                       -0.541202   0.286335  -1.890 0.058745 .  
TOA_381                        0.331605   0.066883   4.958 7.12e-07 ***
TOA_401                       -0.107477   0.175968  -0.611 0.541348    
TOA_421                        1.142201   0.441189   2.589 0.009628 ** 
TOA_441                        2.153505   1.071909   2.009 0.044533 *  
PPGROUP_111                   -0.408061   0.116673  -3.497 0.000470 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 20683  on 14919  degrees of freedom
Residual deviance: 14545  on 14859  degrees of freedom
AIC: 14667

Number of Fisher Scoring iterations: 5

                                   VIF
GSEGRD                        1.512280
LowerLimitAge                 2.251018
BLS_FEDERAL_OtherSep_Rate     1.340790
BLS_FEDERAL_Quits_Rate        1.211930
BLS_FEDERAL_JobOpenings_Level 1.089395
LOSSqrt                       1.730883
SEPCount_EFDATE_OCCLog        1.102851
AGELVL_B                      1.164450
AGELVL_C                      1.660398
AGELVL_E                      1.404859
AGELVL_F                      1.279586
AGELVL_G                      1.266799
AGELVL_H                      1.357462
AGELVL_I                      1.513758
LOC_01                        1.053598
LOC_02                        1.030629
LOC_04                        1.108536
LOC_06                        1.222225
LOC_08                        1.100523
LOC_13                        1.120631
LOC_15                        1.065337
LOC_17                        1.069896
LOC_18                        1.028260
LOC_19                        1.017193
LOC_20                        1.026470
LOC_22                        1.037291
LOC_24                        1.265115
LOC_25                        1.042141
LOC_26                        1.042619
LOC_27                        1.030876
LOC_28                        1.026272
LOC_29                        1.057965
LOC_30                        1.036981
LOC_32                        1.030589
LOC_34                        1.042095
LOC_35                        1.086186
LOC_38                        1.015875
LOC_39                        1.076969
LOC_40                        1.059639
LOC_41                        1.045000
LOC_42                        1.074458
LOC_44                        1.007978
LOC_46                        1.035415
LOC_47                        1.036268
LOC_48                        1.214284
LOC_49                        1.036447
LOC_50                        1.005599
LOC_51                        1.290264
LOC_53                        1.101776
LOC_55                        1.030730
TOA_15                        1.403390
TOA_20                        1.027235
TOA_30                        1.217459
TOA_32                        1.005620
TOA_35                        1.075770
TOA_38                        1.162608
TOA_40                        1.056915
TOA_42                        1.009867
TOA_44                        1.003207
PPGROUP_11                    1.148730

Removed AFTER this step: LOC_26 


Call:
glm(formula = fla, family = "binomial", data = OPMAnalysisDataNoFamBinary)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-2.8217  -0.7400  -0.1370   0.8186   3.2318  

Coefficients:
                               Estimate Std. Error z value Pr(>|z|)    
(Intercept)                    2.177785   0.334823   6.504 7.81e-11 ***
GSEGRD                        -0.065336   0.014492  -4.508 6.54e-06 ***
LowerLimitAge                 -0.044694   0.002617 -17.075  < 2e-16 ***
BLS_FEDERAL_OtherSep_Rate      3.293718   0.292885  11.246  < 2e-16 ***
BLS_FEDERAL_Quits_Rate         0.460713   0.312275   1.475 0.140120    
BLS_FEDERAL_JobOpenings_Level  0.017215   0.001848   9.317  < 2e-16 ***
LOSSqrt                       -0.779992   0.022102 -35.291  < 2e-16 ***
SEPCount_EFDATE_OCCLog        -0.031172   0.016061  -1.941 0.052270 .  
AGELVL_B1                     -0.583861   0.219563  -2.659 0.007833 ** 
AGELVL_C1                     -0.166535   0.085287  -1.953 0.050862 .  
AGELVL_E1                      0.275792   0.067253   4.101 4.12e-05 ***
AGELVL_F1                      0.491606   0.070012   7.022 2.19e-12 ***
AGELVL_G1                      0.714010   0.073873   9.665  < 2e-16 ***
AGELVL_H1                      0.848356   0.076150  11.141  < 2e-16 ***
AGELVL_I1                      0.719471   0.081034   8.879  < 2e-16 ***
LOC_011                       -0.225805   0.166476  -1.356 0.174978    
LOC_021                        0.145017   0.250089   0.580 0.562009    
LOC_041                        0.509115   0.135622   3.754 0.000174 ***
LOC_061                        0.403704   0.086312   4.677 2.91e-06 ***
LOC_081                        0.130881   0.124501   1.051 0.293148    
LOC_131                        0.154271   0.114273   1.350 0.177010    
LOC_151                       -0.079057   0.159731  -0.495 0.620644    
LOC_171                       -0.233109   0.143310  -1.627 0.103821    
LOC_181                       -0.242535   0.249656  -0.971 0.331311    
LOC_191                        0.199728   0.318453   0.627 0.530541    
LOC_201                        0.597655   0.254514   2.348 0.018863 *  
LOC_221                       -0.126179   0.206280  -0.612 0.540744    
LOC_241                       -0.119141   0.081869  -1.455 0.145598    
LOC_251                        0.134897   0.180310   0.748 0.454376    
LOC_271                        0.274773   0.236268   1.163 0.244841    
LOC_281                       -0.141487   0.244446  -0.579 0.562720    
LOC_291                       -0.365224   0.167620  -2.179 0.029340 *  
LOC_301                        0.665861   0.238616   2.791 0.005262 ** 
LOC_321                        0.215979   0.242419   0.891 0.372966    
LOC_341                       -0.190084   0.194105  -0.979 0.327440    
LOC_351                        0.455090   0.149481   3.044 0.002331 ** 
LOC_381                        0.472262   0.365556   1.292 0.196393    
LOC_391                       -0.303906   0.141126  -2.153 0.031284 *  
LOC_401                        0.163045   0.164110   0.994 0.320464    
LOC_411                        0.100774   0.195393   0.516 0.606028    
LOC_421                       -0.308857   0.142147  -2.173 0.029795 *  
LOC_441                        0.326790   0.490707   0.666 0.505439    
LOC_461                        0.797511   0.252398   3.160 0.001579 ** 
LOC_471                       -0.196007   0.208316  -0.941 0.346751    
LOC_481                        0.238809   0.091221   2.618 0.008847 ** 
LOC_491                       -0.200400   0.207233  -0.967 0.333530    
LOC_501                       -0.430106   0.591634  -0.727 0.467239    
LOC_511                       -0.166400   0.077517  -2.147 0.031822 *  
LOC_531                        0.385956   0.127067   3.037 0.002386 ** 
LOC_551                       -0.335025   0.249030  -1.345 0.178522    
TOA_151                        0.067252   0.069041   0.974 0.330012    
TOA_201                        1.125435   0.192715   5.840 5.22e-09 ***
TOA_301                        0.342393   0.088124   3.885 0.000102 ***
TOA_321                        0.417338   0.754246   0.553 0.580046    
TOA_351                       -0.545449   0.286238  -1.906 0.056705 .  
TOA_381                        0.330550   0.066833   4.946 7.58e-07 ***
TOA_401                       -0.107444   0.175955  -0.611 0.541443    
TOA_421                        1.143245   0.441175   2.591 0.009560 ** 
TOA_441                        2.156214   1.071893   2.012 0.044263 *  
PPGROUP_111                   -0.408148   0.116678  -3.498 0.000469 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 20683  on 14919  degrees of freedom
Residual deviance: 14545  on 14860  degrees of freedom
AIC: 14665

Number of Fisher Scoring iterations: 5

                                   VIF
GSEGRD                        1.506995
LowerLimitAge                 2.250918
BLS_FEDERAL_OtherSep_Rate     1.340706
BLS_FEDERAL_Quits_Rate        1.211608
BLS_FEDERAL_JobOpenings_Level 1.089375
LOSSqrt                       1.730760
SEPCount_EFDATE_OCCLog        1.102818
AGELVL_B                      1.164402
AGELVL_C                      1.660390
AGELVL_E                      1.404808
AGELVL_F                      1.279548
AGELVL_G                      1.266786
AGELVL_H                      1.357452
AGELVL_I                      1.513875
LOC_01                        1.050524
LOC_02                        1.029147
LOC_04                        1.103224
LOC_06                        1.208461
LOC_08                        1.094939
LOC_13                        1.114088
LOC_15                        1.062019
LOC_17                        1.065725
LOC_18                        1.026702
LOC_19                        1.016233
LOC_20                        1.025059
LOC_22                        1.035150
LOC_24                        1.253144
LOC_25                        1.039673
LOC_27                        1.029097
LOC_28                        1.024737
LOC_29                        1.054584
LOC_30                        1.035252
LOC_32                        1.028939
LOC_34                        1.039802
LOC_35                        1.081890
LOC_38                        1.015156
LOC_39                        1.072350
LOC_40                        1.056385
LOC_41                        1.042503
LOC_42                        1.070213
LOC_44                        1.007605
LOC_46                        1.033782
LOC_47                        1.034125
LOC_48                        1.202093
LOC_49                        1.034435
LOC_50                        1.005344
LOC_51                        1.274819
LOC_53                        1.096022
LOC_55                        1.029062
TOA_15                        1.403215
TOA_20                        1.027209
TOA_30                        1.217094
TOA_32                        1.005618
TOA_35                        1.074343
TOA_38                        1.161137
TOA_40                        1.056920
TOA_42                        1.009835
TOA_44                        1.003174
PPGROUP_11                    1.148692

Removed AFTER this step: LOC_15 


Call:
glm(formula = fla, family = "binomial", data = OPMAnalysisDataNoFamBinary)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-2.8213  -0.7403  -0.1371   0.8185   3.2324  

Coefficients:
                               Estimate Std. Error z value Pr(>|z|)    
(Intercept)                    2.167100   0.334112   6.486 8.81e-11 ***
GSEGRD                        -0.064988   0.014474  -4.490 7.12e-06 ***
LowerLimitAge                 -0.044706   0.002617 -17.081  < 2e-16 ***
BLS_FEDERAL_OtherSep_Rate      3.295427   0.292864  11.252  < 2e-16 ***
BLS_FEDERAL_Quits_Rate         0.460022   0.312275   1.473 0.140715    
BLS_FEDERAL_JobOpenings_Level  0.017221   0.001848   9.321  < 2e-16 ***
LOSSqrt                       -0.779825   0.022098 -35.290  < 2e-16 ***
SEPCount_EFDATE_OCCLog        -0.031434   0.016053  -1.958 0.050212 .  
AGELVL_B1                     -0.582016   0.219528  -2.651 0.008020 ** 
AGELVL_C1                     -0.166465   0.085283  -1.952 0.050948 .  
AGELVL_E1                      0.275444   0.067249   4.096 4.21e-05 ***
AGELVL_F1                      0.491826   0.070009   7.025 2.14e-12 ***
AGELVL_G1                      0.713842   0.073876   9.663  < 2e-16 ***
AGELVL_H1                      0.848003   0.076146  11.136  < 2e-16 ***
AGELVL_I1                      0.718703   0.081016   8.871  < 2e-16 ***
LOC_011                       -0.220532   0.166130  -1.327 0.184354    
LOC_021                        0.150877   0.249787   0.604 0.545827    
LOC_041                        0.515111   0.135076   3.813 0.000137 ***
LOC_061                        0.409032   0.085633   4.777 1.78e-06 ***
LOC_081                        0.136450   0.123983   1.101 0.271090    
LOC_131                        0.159834   0.113715   1.406 0.159850    
LOC_171                       -0.227807   0.142909  -1.594 0.110919    
LOC_181                       -0.237452   0.249441  -0.952 0.341129    
LOC_191                        0.204964   0.318292   0.644 0.519609    
LOC_201                        0.603451   0.254227   2.374 0.017613 *  
LOC_221                       -0.120642   0.205971  -0.586 0.558060    
LOC_241                       -0.113956   0.081194  -1.404 0.160463    
LOC_251                        0.140003   0.180012   0.778 0.436721    
LOC_271                        0.280256   0.236002   1.188 0.235025    
LOC_281                       -0.136051   0.244189  -0.557 0.577423    
LOC_291                       -0.359645   0.167232  -2.151 0.031510 *  
LOC_301                        0.671827   0.238296   2.819 0.004813 ** 
LOC_321                        0.221209   0.242185   0.913 0.361039    
LOC_341                       -0.184926   0.193819  -0.954 0.340026    
LOC_351                        0.461018   0.148993   3.094 0.001973 ** 
LOC_381                        0.478059   0.365342   1.309 0.190696    
LOC_391                       -0.298616   0.140718  -2.122 0.033830 *  
LOC_401                        0.168755   0.163701   1.031 0.302600    
LOC_411                        0.106113   0.195093   0.544 0.586503    
LOC_421                       -0.303676   0.141758  -2.142 0.032177 *  
LOC_441                        0.332141   0.490624   0.677 0.498420    
LOC_461                        0.803656   0.252079   3.188 0.001432 ** 
LOC_471                       -0.190606   0.208029  -0.916 0.359538    
LOC_481                        0.244468   0.090497   2.701 0.006905 ** 
LOC_491                       -0.194813   0.206919  -0.941 0.346453    
LOC_501                       -0.424753   0.591543  -0.718 0.472731    
LOC_511                       -0.160960   0.076726  -2.098 0.035918 *  
LOC_531                        0.391562   0.126556   3.094 0.001975 ** 
LOC_551                       -0.329849   0.248806  -1.326 0.184929    
TOA_151                        0.067198   0.069042   0.973 0.330409    
TOA_201                        1.125152   0.192690   5.839 5.25e-09 ***
TOA_301                        0.344818   0.087979   3.919 8.88e-05 ***
TOA_321                        0.416961   0.754295   0.553 0.580413    
TOA_351                       -0.542589   0.286165  -1.896 0.057951 .  
TOA_381                        0.331979   0.066769   4.972 6.62e-07 ***
TOA_401                       -0.105069   0.175882  -0.597 0.550252    
TOA_421                        1.144330   0.441190   2.594 0.009494 ** 
TOA_441                        2.160675   1.071851   2.016 0.043817 *  
PPGROUP_111                   -0.406781   0.116648  -3.487 0.000488 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 20683  on 14919  degrees of freedom
Residual deviance: 14545  on 14861  degrees of freedom
AIC: 14663

Number of Fisher Scoring iterations: 5

                                   VIF
GSEGRD                        1.503479
LowerLimitAge                 2.250805
BLS_FEDERAL_OtherSep_Rate     1.340598
BLS_FEDERAL_Quits_Rate        1.211633
BLS_FEDERAL_JobOpenings_Level 1.089331
LOSSqrt                       1.730392
SEPCount_EFDATE_OCCLog        1.101626
AGELVL_B                      1.163997
AGELVL_C                      1.660373
AGELVL_E                      1.404599
AGELVL_F                      1.279487
AGELVL_G                      1.266678
AGELVL_H                      1.357266
AGELVL_I                      1.513259
LOC_01                        1.046210
LOC_02                        1.026840
LOC_04                        1.094413
LOC_06                        1.189597
LOC_08                        1.085976
LOC_13                        1.103284
LOC_17                        1.059736
LOC_18                        1.024950
LOC_19                        1.015103
LOC_20                        1.022888
LOC_22                        1.032090
LOC_24                        1.232548
LOC_25                        1.036251
LOC_27                        1.026820
LOC_28                        1.022660
LOC_29                        1.049793
LOC_30                        1.032611
LOC_32                        1.026972
LOC_34                        1.036780
LOC_35                        1.074937
LOC_38                        1.014109
LOC_39                        1.066168
LOC_40                        1.051144
LOC_41                        1.039316
LOC_42                        1.064367
LOC_44                        1.007114
LOC_46                        1.031278
LOC_47                        1.031270
LOC_48                        1.183171
LOC_49                        1.031354
LOC_50                        1.005006
LOC_51                        1.249082
LOC_53                        1.087281
LOC_55                        1.027229
TOA_15                        1.403272
TOA_20                        1.027213
TOA_30                        1.213379
TOA_32                        1.005617
TOA_35                        1.073867
TOA_38                        1.158998
TOA_40                        1.056125
TOA_42                        1.009808
TOA_44                        1.003102
PPGROUP_11                    1.148037

Removed AFTER this step: LOC_41 


Call:
glm(formula = fla, family = "binomial", data = OPMAnalysisDataNoFamBinary)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-2.8220  -0.7401  -0.1371   0.8186   3.2315  

Coefficients:
                               Estimate Std. Error z value Pr(>|z|)    
(Intercept)                    2.178870   0.333420   6.535 6.36e-11 ***
GSEGRD                        -0.065464   0.014448  -4.531 5.87e-06 ***
LowerLimitAge                 -0.044689   0.002617 -17.077  < 2e-16 ***
BLS_FEDERAL_OtherSep_Rate      3.294949   0.292863  11.251  < 2e-16 ***
BLS_FEDERAL_Quits_Rate         0.460686   0.312265   1.475 0.140131    
BLS_FEDERAL_JobOpenings_Level  0.017222   0.001848   9.322  < 2e-16 ***
LOSSqrt                       -0.779652   0.022096 -35.285  < 2e-16 ***
SEPCount_EFDATE_OCCLog        -0.031881   0.016033  -1.988 0.046767 *  
AGELVL_B1                     -0.582450   0.219560  -2.653 0.007983 ** 
AGELVL_C1                     -0.167140   0.085275  -1.960 0.049995 *  
AGELVL_E1                      0.275276   0.067249   4.093 4.25e-05 ***
AGELVL_F1                      0.492073   0.070005   7.029 2.08e-12 ***
AGELVL_G1                      0.713980   0.073875   9.665  < 2e-16 ***
AGELVL_H1                      0.848069   0.076138  11.139  < 2e-16 ***
AGELVL_I1                      0.719032   0.081014   8.875  < 2e-16 ***
LOC_011                       -0.225058   0.165919  -1.356 0.174963    
LOC_021                        0.145821   0.249603   0.584 0.559077    
LOC_041                        0.510348   0.134785   3.786 0.000153 ***
LOC_061                        0.404403   0.085207   4.746 2.07e-06 ***
LOC_081                        0.131928   0.123700   1.067 0.286191    
LOC_131                        0.155365   0.113413   1.370 0.170716    
LOC_171                       -0.232225   0.142671  -1.628 0.103590    
LOC_181                       -0.242055   0.249284  -0.971 0.331549    
LOC_191                        0.199885   0.318132   0.628 0.529802    
LOC_201                        0.598674   0.254054   2.356 0.018449 *  
LOC_221                       -0.125437   0.205772  -0.610 0.542131    
LOC_241                       -0.118063   0.080841  -1.460 0.144168    
LOC_251                        0.135587   0.179818   0.754 0.450835    
LOC_271                        0.275300   0.235823   1.167 0.243049    
LOC_281                       -0.140905   0.244009  -0.577 0.563631    
LOC_291                       -0.364353   0.167004  -2.182 0.029132 *  
LOC_301                        0.666626   0.238084   2.800 0.005111 ** 
LOC_321                        0.216197   0.241998   0.893 0.371652    
LOC_341                       -0.189343   0.193646  -0.978 0.328182    
LOC_351                        0.456394   0.148745   3.068 0.002153 ** 
LOC_381                        0.472684   0.365169   1.294 0.195518    
LOC_391                       -0.303184   0.140464  -2.158 0.030893 *  
LOC_401                        0.164191   0.163476   1.004 0.315197    
LOC_421                       -0.307952   0.141536  -2.176 0.029572 *  
LOC_441                        0.327239   0.490482   0.667 0.504659    
LOC_461                        0.798665   0.251894   3.171 0.001521 ** 
LOC_471                       -0.195362   0.207841  -0.940 0.347238    
LOC_481                        0.239898   0.090104   2.662 0.007757 ** 
LOC_491                       -0.199520   0.206729  -0.965 0.334480    
LOC_501                       -0.429365   0.591445  -0.726 0.467864    
LOC_511                       -0.165273   0.076318  -2.166 0.030344 *  
LOC_531                        0.386865   0.126254   3.064 0.002183 ** 
LOC_551                       -0.334747   0.248628  -1.346 0.178182    
TOA_151                        0.066742   0.069034   0.967 0.333643    
TOA_201                        1.124479   0.192695   5.836 5.36e-09 ***
TOA_301                        0.344001   0.087958   3.911 9.19e-05 ***
TOA_321                        0.417067   0.754190   0.553 0.580263    
TOA_351                       -0.544156   0.286159  -1.902 0.057225 .  
TOA_381                        0.332126   0.066773   4.974 6.56e-07 ***
TOA_401                       -0.106075   0.175879  -0.603 0.546433    
TOA_421                        1.145463   0.441146   2.597 0.009416 ** 
TOA_441                        2.157406   1.071822   2.013 0.044131 *  
PPGROUP_111                   -0.407702   0.116632  -3.496 0.000473 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 20683  on 14919  degrees of freedom
Residual deviance: 14545  on 14862  degrees of freedom
AIC: 14661

Number of Fisher Scoring iterations: 5

                                   VIF
GSEGRD                        1.497918
LowerLimitAge                 2.250272
BLS_FEDERAL_OtherSep_Rate     1.340597
BLS_FEDERAL_Quits_Rate        1.211599
BLS_FEDERAL_JobOpenings_Level 1.089337
LOSSqrt                       1.729903
SEPCount_EFDATE_OCCLog        1.098816
AGELVL_B                      1.164069
AGELVL_C                      1.660025
AGELVL_E                      1.404572
AGELVL_F                      1.279456
AGELVL_G                      1.266619
AGELVL_H                      1.357234
AGELVL_I                      1.512974
LOC_01                        1.043598
LOC_02                        1.025422
LOC_04                        1.089839
LOC_06                        1.177903
LOC_08                        1.081119
LOC_13                        1.097553
LOC_17                        1.056338
LOC_18                        1.023783
LOC_19                        1.014235
LOC_20                        1.021675
LOC_22                        1.030213
LOC_24                        1.221963
LOC_25                        1.034158
LOC_27                        1.025298
LOC_28                        1.021304
LOC_29                        1.047001
LOC_30                        1.030953
LOC_32                        1.025495
LOC_34                        1.034977
LOC_35                        1.071456
LOC_38                        1.013371
LOC_39                        1.062398
LOC_40                        1.048404
LOC_42                        1.061124
LOC_44                        1.006775
LOC_46                        1.029916
LOC_47                        1.029458
LOC_48                        1.173022
LOC_49                        1.029562
LOC_50                        1.004802
LOC_51                        1.235817
LOC_53                        1.082251
LOC_55                        1.025900
TOA_15                        1.402979
TOA_20                        1.027189
TOA_30                        1.212992
TOA_32                        1.005617
TOA_35                        1.073894
TOA_38                        1.158938
TOA_40                        1.056016
TOA_42                        1.009783
TOA_44                        1.003071
PPGROUP_11                    1.147824

Removed AFTER this step: TOA_32 


Call:
glm(formula = fla, family = "binomial", data = OPMAnalysisDataNoFamBinary)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-2.8217  -0.7405  -0.1372   0.8185   3.2315  

Coefficients:
                               Estimate Std. Error z value Pr(>|z|)    
(Intercept)                    2.176247   0.333377   6.528 6.67e-11 ***
GSEGRD                        -0.065233   0.014441  -4.517 6.27e-06 ***
LowerLimitAge                 -0.044682   0.002617 -17.074  < 2e-16 ***
BLS_FEDERAL_OtherSep_Rate      3.295024   0.292860  11.251  < 2e-16 ***
BLS_FEDERAL_Quits_Rate         0.461875   0.312251   1.479 0.139092    
BLS_FEDERAL_JobOpenings_Level  0.017209   0.001847   9.315  < 2e-16 ***
LOSSqrt                       -0.779712   0.022095 -35.288  < 2e-16 ***
SEPCount_EFDATE_OCCLog        -0.031981   0.016032  -1.995 0.046059 *  
AGELVL_B1                     -0.582063   0.219556  -2.651 0.008023 ** 
AGELVL_C1                     -0.167290   0.085273  -1.962 0.049784 *  
AGELVL_E1                      0.275584   0.067247   4.098 4.17e-05 ***
AGELVL_F1                      0.491817   0.070001   7.026 2.13e-12 ***
AGELVL_G1                      0.714831   0.073860   9.678  < 2e-16 ***
AGELVL_H1                      0.847644   0.076133  11.134  < 2e-16 ***
AGELVL_I1                      0.718729   0.081013   8.872  < 2e-16 ***
LOC_011                       -0.225053   0.165912  -1.356 0.174951    
LOC_021                        0.145989   0.249587   0.585 0.558601    
LOC_041                        0.510591   0.134781   3.788 0.000152 ***
LOC_061                        0.405592   0.085180   4.762 1.92e-06 ***
LOC_081                        0.131946   0.123694   1.067 0.286101    
LOC_131                        0.155377   0.113409   1.370 0.170669    
LOC_171                       -0.232178   0.142666  -1.627 0.103648    
LOC_181                       -0.241865   0.249271  -0.970 0.331902    
LOC_191                        0.199976   0.318124   0.629 0.529604    
LOC_201                        0.598850   0.254046   2.357 0.018411 *  
LOC_221                       -0.125316   0.205764  -0.609 0.542506    
LOC_241                       -0.118207   0.080837  -1.462 0.143666    
LOC_251                        0.135651   0.179811   0.754 0.450602    
LOC_271                        0.275515   0.235814   1.168 0.242662    
LOC_281                       -0.139292   0.244122  -0.571 0.568282    
LOC_291                       -0.364184   0.166997  -2.181 0.029199 *  
LOC_301                        0.666838   0.238071   2.801 0.005094 ** 
LOC_321                        0.216301   0.241985   0.894 0.371397    
LOC_341                       -0.189289   0.193641  -0.978 0.328310    
LOC_351                        0.456627   0.148738   3.070 0.002141 ** 
LOC_381                        0.472784   0.365140   1.295 0.195389    
LOC_391                       -0.301911   0.140495  -2.149 0.031641 *  
LOC_401                        0.164374   0.163470   1.006 0.314643    
LOC_421                       -0.307922   0.141533  -2.176 0.029583 *  
LOC_441                        0.327257   0.490465   0.667 0.504620    
LOC_461                        0.799018   0.251882   3.172 0.001513 ** 
LOC_471                       -0.195241   0.207835  -0.939 0.347523    
LOC_481                        0.239965   0.090102   2.663 0.007739 ** 
LOC_491                       -0.199426   0.206719  -0.965 0.334687    
LOC_501                       -0.429463   0.591412  -0.726 0.467737    
LOC_511                       -0.163443   0.076245  -2.144 0.032061 *  
LOC_531                        0.386944   0.126251   3.065 0.002178 ** 
LOC_551                       -0.334569   0.248620  -1.346 0.178397    
TOA_151                        0.066393   0.069031   0.962 0.336158    
TOA_201                        1.124169   0.192691   5.834 5.41e-09 ***
TOA_301                        0.343255   0.087945   3.903 9.50e-05 ***
TOA_351                       -0.543946   0.286153  -1.901 0.057316 .  
TOA_381                        0.331762   0.066770   4.969 6.74e-07 ***
TOA_401                       -0.106546   0.175870  -0.606 0.544633    
TOA_421                        1.144861   0.441131   2.595 0.009451 ** 
TOA_441                        2.156808   1.071823   2.012 0.044190 *  
PPGROUP_111                   -0.407082   0.116626  -3.490 0.000482 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 20683  on 14919  degrees of freedom
Residual deviance: 14546  on 14863  degrees of freedom
AIC: 14660

Number of Fisher Scoring iterations: 5

                                   VIF
GSEGRD                        1.496766
LowerLimitAge                 2.250294
BLS_FEDERAL_OtherSep_Rate     1.340601
BLS_FEDERAL_Quits_Rate        1.211541
BLS_FEDERAL_JobOpenings_Level 1.089152
LOSSqrt                       1.729923
SEPCount_EFDATE_OCCLog        1.098668
AGELVL_B                      1.164040
AGELVL_C                      1.660037
AGELVL_E                      1.404476
AGELVL_F                      1.279409
AGELVL_G                      1.266206
AGELVL_H                      1.357132
AGELVL_I                      1.512955
LOC_01                        1.043600
LOC_02                        1.025423
LOC_04                        1.089832
LOC_06                        1.177227
LOC_08                        1.081124
LOC_13                        1.097555
LOC_17                        1.056339
LOC_18                        1.023781
LOC_19                        1.014235
LOC_20                        1.021674
LOC_22                        1.030213
LOC_24                        1.221957
LOC_25                        1.034159
LOC_27                        1.025297
LOC_28                        1.021065
LOC_29                        1.046999
LOC_30                        1.030954
LOC_32                        1.025496
LOC_34                        1.034975
LOC_35                        1.071452
LOC_38                        1.013372
LOC_39                        1.061973
LOC_40                        1.048400
LOC_42                        1.061125
LOC_44                        1.006775
LOC_46                        1.029913
LOC_47                        1.029457
LOC_48                        1.173022
LOC_49                        1.029563
LOC_50                        1.004802
LOC_51                        1.233545
LOC_53                        1.082250
LOC_55                        1.025898
TOA_15                        1.402889
TOA_20                        1.027183
TOA_30                        1.212698
TOA_35                        1.073886
TOA_38                        1.158843
TOA_40                        1.056003
TOA_42                        1.009778
TOA_44                        1.003070
PPGROUP_11                    1.147716

Removed AFTER this step: LOC_28 


Call:
glm(formula = fla, family = "binomial", data = OPMAnalysisDataNoFamBinary)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-2.8215  -0.7405  -0.1380   0.8185   3.2322  

Coefficients:
                               Estimate Std. Error z value Pr(>|z|)    
(Intercept)                    2.169054   0.333129   6.511 7.46e-11 ***
GSEGRD                        -0.064945   0.014431  -4.500 6.79e-06 ***
LowerLimitAge                 -0.044685   0.002617 -17.075  < 2e-16 ***
BLS_FEDERAL_OtherSep_Rate      3.294316   0.292857  11.249  < 2e-16 ***
BLS_FEDERAL_Quits_Rate         0.462407   0.312235   1.481 0.138617    
BLS_FEDERAL_JobOpenings_Level  0.017205   0.001847   9.313  < 2e-16 ***
LOSSqrt                       -0.779818   0.022095 -35.293  < 2e-16 ***
SEPCount_EFDATE_OCCLog        -0.031867   0.016030  -1.988 0.046810 *  
AGELVL_B1                     -0.580881   0.219532  -2.646 0.008145 ** 
AGELVL_C1                     -0.166532   0.085262  -1.953 0.050798 .  
AGELVL_E1                      0.276004   0.067240   4.105 4.05e-05 ***
AGELVL_F1                      0.492085   0.069995   7.030 2.06e-12 ***
AGELVL_G1                      0.715377   0.073857   9.686  < 2e-16 ***
AGELVL_H1                      0.847943   0.076130  11.138  < 2e-16 ***
AGELVL_I1                      0.718427   0.081014   8.868  < 2e-16 ***
LOC_011                       -0.221452   0.165793  -1.336 0.181642    
LOC_021                        0.149896   0.249494   0.601 0.547974    
LOC_041                        0.514401   0.134620   3.821 0.000133 ***
LOC_061                        0.409229   0.084942   4.818 1.45e-06 ***
LOC_081                        0.135539   0.123535   1.097 0.272567    
LOC_131                        0.158946   0.113238   1.404 0.160427    
LOC_171                       -0.228645   0.142535  -1.604 0.108684    
LOC_181                       -0.238148   0.249191  -0.956 0.339231    
LOC_191                        0.203862   0.318063   0.641 0.521556    
LOC_201                        0.602653   0.253968   2.373 0.017647 *  
LOC_221                       -0.121581   0.205665  -0.591 0.554412    
LOC_241                       -0.114905   0.080629  -1.425 0.154128    
LOC_251                        0.139141   0.179713   0.774 0.438786    
LOC_271                        0.279380   0.235715   1.185 0.235922    
LOC_291                       -0.360457   0.166870  -2.160 0.030764 *  
LOC_301                        0.670852   0.237976   2.819 0.004818 ** 
LOC_321                        0.220143   0.241894   0.910 0.362780    
LOC_341                       -0.185793   0.193545  -0.960 0.337080    
LOC_351                        0.460377   0.148596   3.098 0.001947 ** 
LOC_381                        0.476738   0.365090   1.306 0.191617    
LOC_391                       -0.298290   0.140353  -2.125 0.033563 *  
LOC_401                        0.168039   0.163349   1.029 0.303616    
LOC_421                       -0.304488   0.141407  -2.153 0.031298 *  
LOC_441                        0.331054   0.490464   0.675 0.499688    
LOC_461                        0.802992   0.251793   3.189 0.001427 ** 
LOC_471                       -0.191499   0.207734  -0.922 0.356608    
LOC_481                        0.243618   0.089875   2.711 0.006716 ** 
LOC_491                       -0.195699   0.206621  -0.947 0.343567    
LOC_501                       -0.425743   0.591387  -0.720 0.471583    
LOC_511                       -0.159954   0.075996  -2.105 0.035313 *  
LOC_531                        0.390625   0.126090   3.098 0.001948 ** 
LOC_551                       -0.330756   0.248533  -1.331 0.183243    
TOA_151                        0.066255   0.069034   0.960 0.337180    
TOA_201                        1.125226   0.192682   5.840 5.23e-09 ***
TOA_301                        0.344208   0.087937   3.914 9.07e-05 ***
TOA_351                       -0.543791   0.286142  -1.900 0.057378 .  
TOA_381                        0.331520   0.066769   4.965 6.86e-07 ***
TOA_401                       -0.105577   0.175858  -0.600 0.548270    
TOA_421                        1.146256   0.441115   2.599 0.009362 ** 
TOA_441                        2.159517   1.071813   2.015 0.043923 *  
PPGROUP_111                   -0.406854   0.116629  -3.488 0.000486 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 20683  on 14919  degrees of freedom
Residual deviance: 14546  on 14864  degrees of freedom
AIC: 14658

Number of Fisher Scoring iterations: 5

                                   VIF
GSEGRD                        1.494861
LowerLimitAge                 2.250410
BLS_FEDERAL_OtherSep_Rate     1.340539
BLS_FEDERAL_Quits_Rate        1.211511
BLS_FEDERAL_JobOpenings_Level 1.089136
LOSSqrt                       1.729950
SEPCount_EFDATE_OCCLog        1.098513
AGELVL_B                      1.163762
AGELVL_C                      1.659602
AGELVL_E                      1.404343
AGELVL_F                      1.279406
AGELVL_G                      1.266026
AGELVL_H                      1.357184
AGELVL_I                      1.512943
LOC_01                        1.042083
LOC_02                        1.024650
LOC_04                        1.087148
LOC_06                        1.170599
LOC_08                        1.078309
LOC_13                        1.094198
LOC_17                        1.054332
LOC_18                        1.023077
LOC_19                        1.013767
LOC_20                        1.020968
LOC_22                        1.029161
LOC_24                        1.215632
LOC_25                        1.032952
LOC_27                        1.024447
LOC_29                        1.045386
LOC_30                        1.030052
LOC_32                        1.024698
LOC_34                        1.033927
LOC_35                        1.069356
LOC_38                        1.013003
LOC_39                        1.059792
LOC_40                        1.046769
LOC_42                        1.059184
LOC_44                        1.006589
LOC_46                        1.029125
LOC_47                        1.028426
LOC_48                        1.167075
LOC_49                        1.028528
LOC_50                        1.004679
LOC_51                        1.225560
LOC_53                        1.079402
LOC_55                        1.025149
TOA_15                        1.402840
TOA_20                        1.027086
TOA_30                        1.212261
TOA_35                        1.073691
TOA_38                        1.158770
TOA_40                        1.055898
TOA_42                        1.009745
TOA_44                        1.003050
PPGROUP_11                    1.147690

Removed AFTER this step: LOC_22 


Call:
glm(formula = fla, family = "binomial", data = OPMAnalysisDataNoFamBinary)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-2.8203  -0.7409  -0.1380   0.8186   3.2332  

Coefficients:
                               Estimate Std. Error z value Pr(>|z|)    
(Intercept)                    2.161288   0.332870   6.493 8.42e-11 ***
GSEGRD                        -0.064562   0.014417  -4.478 7.52e-06 ***
LowerLimitAge                 -0.044713   0.002617 -17.088  < 2e-16 ***
BLS_FEDERAL_OtherSep_Rate      3.294886   0.292855  11.251  < 2e-16 ***
BLS_FEDERAL_Quits_Rate         0.460607   0.312216   1.475 0.140136    
BLS_FEDERAL_JobOpenings_Level  0.017199   0.001847   9.310  < 2e-16 ***
LOSSqrt                       -0.779801   0.022096 -35.292  < 2e-16 ***
SEPCount_EFDATE_OCCLog        -0.031724   0.016029  -1.979 0.047788 *  
AGELVL_B1                     -0.579085   0.219504  -2.638 0.008336 ** 
AGELVL_C1                     -0.166917   0.085256  -1.958 0.050251 .  
AGELVL_E1                      0.275786   0.067237   4.102 4.10e-05 ***
AGELVL_F1                      0.492026   0.070000   7.029 2.08e-12 ***
AGELVL_G1                      0.715670   0.073853   9.690  < 2e-16 ***
AGELVL_H1                      0.848281   0.076129  11.143  < 2e-16 ***
AGELVL_I1                      0.718516   0.081011   8.869  < 2e-16 ***
LOC_011                       -0.217237   0.165633  -1.312 0.189670    
LOC_021                        0.154550   0.249355   0.620 0.535390    
LOC_041                        0.518977   0.134396   3.862 0.000113 ***
LOC_061                        0.413593   0.084618   4.888 1.02e-06 ***
LOC_081                        0.139877   0.123311   1.134 0.256652    
LOC_131                        0.163207   0.113006   1.444 0.148672    
LOC_171                       -0.224452   0.142355  -1.577 0.114863    
LOC_181                       -0.233653   0.249068  -0.938 0.348188    
LOC_191                        0.208435   0.317976   0.656 0.512142    
LOC_201                        0.607172   0.253847   2.392 0.016762 *  
LOC_241                       -0.110925   0.080345  -1.381 0.167397    
LOC_251                        0.143281   0.179572   0.798 0.424928    
LOC_271                        0.283939   0.235582   1.205 0.228099    
LOC_291                       -0.355953   0.166688  -2.135 0.032724 *  
LOC_301                        0.675637   0.237836   2.841 0.004500 ** 
LOC_321                        0.224748   0.241765   0.930 0.352573    
LOC_341                       -0.181566   0.193407  -0.939 0.347844    
LOC_351                        0.464899   0.148396   3.133 0.001731 ** 
LOC_381                        0.481601   0.364993   1.319 0.187008    
LOC_391                       -0.294018   0.140163  -2.098 0.035932 *  
LOC_401                        0.172461   0.163178   1.057 0.290560    
LOC_421                       -0.300337   0.141229  -2.127 0.033453 *  
LOC_441                        0.335543   0.490436   0.684 0.493866    
LOC_461                        0.807800   0.251665   3.210 0.001328 ** 
LOC_471                       -0.187048   0.207590  -0.901 0.367566    
LOC_481                        0.247992   0.089568   2.769 0.005627 ** 
LOC_491                       -0.191196   0.206470  -0.926 0.354433    
LOC_501                       -0.421389   0.591355  -0.713 0.476104    
LOC_511                       -0.155766   0.075658  -2.059 0.039513 *  
LOC_531                        0.395079   0.125863   3.139 0.001695 ** 
LOC_551                       -0.326195   0.248409  -1.313 0.189136    
TOA_151                        0.066241   0.069036   0.960 0.337299    
TOA_201                        1.123629   0.192663   5.832 5.47e-09 ***
TOA_301                        0.345404   0.087910   3.929 8.53e-05 ***
TOA_351                       -0.541986   0.286141  -1.894 0.058209 .  
TOA_381                        0.331559   0.066770   4.966 6.84e-07 ***
TOA_401                       -0.105442   0.175862  -0.600 0.548791    
TOA_421                        1.145778   0.441115   2.597 0.009392 ** 
TOA_441                        2.162627   1.071804   2.018 0.043618 *  
PPGROUP_111                   -0.406502   0.116642  -3.485 0.000492 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 20683  on 14919  degrees of freedom
Residual deviance: 14546  on 14865  degrees of freedom
AIC: 14656

Number of Fisher Scoring iterations: 5

                                   VIF
GSEGRD                        1.491860
LowerLimitAge                 2.249857
BLS_FEDERAL_OtherSep_Rate     1.340624
BLS_FEDERAL_Quits_Rate        1.211398
BLS_FEDERAL_JobOpenings_Level 1.089165
LOSSqrt                       1.730046
SEPCount_EFDATE_OCCLog        1.098293
AGELVL_B                      1.163518
AGELVL_C                      1.659501
AGELVL_E                      1.404328
AGELVL_F                      1.279390
AGELVL_G                      1.266026
AGELVL_H                      1.357187
AGELVL_I                      1.513125
LOC_01                        1.040148
LOC_02                        1.023631
LOC_04                        1.083542
LOC_06                        1.161719
LOC_08                        1.074488
LOC_13                        1.089741
LOC_17                        1.051699
LOC_18                        1.022117
LOC_19                        1.013164
LOC_20                        1.020041
LOC_24                        1.207112
LOC_25                        1.031373
LOC_27                        1.023343
LOC_29                        1.043193
LOC_30                        1.028860
LOC_32                        1.023631
LOC_34                        1.032500
LOC_35                        1.066512
LOC_38                        1.012487
LOC_39                        1.056964
LOC_40                        1.044558
LOC_42                        1.056548
LOC_44                        1.006347
LOC_46                        1.028047
LOC_47                        1.027067
LOC_48                        1.159125
LOC_49                        1.027123
LOC_50                        1.004522
LOC_51                        1.214843
LOC_53                        1.075524
LOC_55                        1.024151
TOA_15                        1.402810
TOA_20                        1.026875
TOA_30                        1.211641
TOA_35                        1.073533
TOA_38                        1.158785
TOA_40                        1.055889
TOA_42                        1.009746
TOA_44                        1.003025
PPGROUP_11                    1.147630

Removed AFTER this step: TOA_40 


Call:
glm(formula = fla, family = "binomial", data = OPMAnalysisDataNoFamBinary)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-2.8205  -0.7409  -0.1383   0.8194   3.1997  

Coefficients:
                               Estimate Std. Error z value Pr(>|z|)    
(Intercept)                    2.159341   0.332844   6.488 8.72e-11 ***
GSEGRD                        -0.065045   0.014394  -4.519 6.22e-06 ***
LowerLimitAge                 -0.044717   0.002617 -17.089  < 2e-16 ***
BLS_FEDERAL_OtherSep_Rate      3.295493   0.292825  11.254  < 2e-16 ***
BLS_FEDERAL_Quits_Rate         0.455954   0.312127   1.461 0.144072    
BLS_FEDERAL_JobOpenings_Level  0.017217   0.001847   9.322  < 2e-16 ***
LOSSqrt                       -0.777807   0.021834 -35.623  < 2e-16 ***
SEPCount_EFDATE_OCCLog        -0.032121   0.016017  -2.005 0.044916 *  
AGELVL_B1                     -0.579293   0.219524  -2.639 0.008319 ** 
AGELVL_C1                     -0.167942   0.085266  -1.970 0.048882 *  
AGELVL_E1                      0.276813   0.067221   4.118 3.82e-05 ***
AGELVL_F1                      0.492899   0.069984   7.043 1.88e-12 ***
AGELVL_G1                      0.716316   0.073835   9.702  < 2e-16 ***
AGELVL_H1                      0.848806   0.076113  11.152  < 2e-16 ***
AGELVL_I1                      0.718648   0.080988   8.874  < 2e-16 ***
LOC_011                       -0.215454   0.165564  -1.301 0.193144    
LOC_021                        0.155552   0.249304   0.624 0.532662    
LOC_041                        0.520412   0.134361   3.873 0.000107 ***
LOC_061                        0.414320   0.084596   4.898 9.70e-07 ***
LOC_081                        0.140973   0.123294   1.143 0.252877    
LOC_131                        0.164903   0.112961   1.460 0.144341    
LOC_171                       -0.223065   0.142328  -1.567 0.117053    
LOC_181                       -0.234853   0.249081  -0.943 0.345744    
LOC_191                        0.209713   0.317886   0.660 0.509439    
LOC_201                        0.607970   0.253788   2.396 0.016594 *  
LOC_241                       -0.112628   0.080334  -1.402 0.160916    
LOC_251                        0.144152   0.179525   0.803 0.421995    
LOC_271                        0.285111   0.235528   1.211 0.226079    
LOC_291                       -0.354741   0.166618  -2.129 0.033248 *  
LOC_301                        0.675529   0.237825   2.840 0.004505 ** 
LOC_321                        0.226042   0.241701   0.935 0.349680    
LOC_341                       -0.179969   0.193374  -0.931 0.352021    
LOC_351                        0.465113   0.148379   3.135 0.001721 ** 
LOC_381                        0.482802   0.364889   1.323 0.185786    
LOC_391                       -0.292731   0.140123  -2.089 0.036699 *  
LOC_401                        0.174402   0.163105   1.069 0.284950    
LOC_421                       -0.300973   0.141256  -2.131 0.033115 *  
LOC_441                        0.330442   0.491305   0.673 0.501214    
LOC_461                        0.809346   0.251594   3.217 0.001296 ** 
LOC_471                       -0.186361   0.207532  -0.898 0.369194    
LOC_481                        0.248828   0.089543   2.779 0.005455 ** 
LOC_491                       -0.189519   0.206412  -0.918 0.358534    
LOC_501                       -0.419348   0.591240  -0.709 0.478158    
LOC_511                       -0.155054   0.075652  -2.050 0.040407 *  
LOC_531                        0.396639   0.125811   3.153 0.001618 ** 
LOC_551                       -0.324583   0.248334  -1.307 0.191198    
TOA_151                        0.071403   0.068487   1.043 0.297144    
TOA_201                        1.127198   0.192524   5.855 4.77e-09 ***
TOA_301                        0.349539   0.087612   3.990 6.62e-05 ***
TOA_351                       -0.536993   0.285998  -1.878 0.060434 .  
TOA_381                        0.334868   0.066521   5.034 4.80e-07 ***
TOA_421                        1.150437   0.441050   2.608 0.009097 ** 
TOA_441                        2.166692   1.071713   2.022 0.043207 *  
PPGROUP_111                   -0.405481   0.116632  -3.477 0.000508 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 20683  on 14919  degrees of freedom
Residual deviance: 14547  on 14866  degrees of freedom
AIC: 14655

Number of Fisher Scoring iterations: 5

                                   VIF
GSEGRD                        1.487185
LowerLimitAge                 2.249881
BLS_FEDERAL_OtherSep_Rate     1.340745
BLS_FEDERAL_Quits_Rate        1.210656
BLS_FEDERAL_JobOpenings_Level 1.088865
LOSSqrt                       1.689806
SEPCount_EFDATE_OCCLog        1.096528
AGELVL_B                      1.163546
AGELVL_C                      1.658576
AGELVL_E                      1.403573
AGELVL_F                      1.278909
AGELVL_G                      1.265762
AGELVL_H                      1.356983
AGELVL_I                      1.513202
LOC_01                        1.039832
LOC_02                        1.023585
LOC_04                        1.083224
LOC_06                        1.161538
LOC_08                        1.074263
LOC_13                        1.089078
LOC_17                        1.051427
LOC_18                        1.022055
LOC_19                        1.013121
LOC_20                        1.020015
LOC_24                        1.205349
LOC_25                        1.031318
LOC_27                        1.023271
LOC_29                        1.043059
LOC_30                        1.028833
LOC_32                        1.023553
LOC_34                        1.032307
LOC_35                        1.066566
LOC_38                        1.012456
LOC_39                        1.056736
LOC_40                        1.044162
LOC_42                        1.056481
LOC_44                        1.006012
LOC_46                        1.027944
LOC_47                        1.027047
LOC_48                        1.158926
LOC_49                        1.026938
LOC_50                        1.004489
LOC_51                        1.214545
LOC_53                        1.075086
LOC_55                        1.024033
TOA_15                        1.380826
TOA_20                        1.025897
TOA_30                        1.204261
TOA_35                        1.072641
TOA_38                        1.150835
TOA_42                        1.009426
TOA_44                        1.002985
PPGROUP_11                    1.147404

Removed AFTER this step: LOC_02 


Call:
glm(formula = fla, family = "binomial", data = OPMAnalysisDataNoFamBinary)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-2.8213  -0.7410  -0.1383   0.8191   3.1986  

Coefficients:
                               Estimate Std. Error z value Pr(>|z|)    
(Intercept)                    2.169022   0.332474   6.524 6.85e-11 ***
GSEGRD                        -0.065595   0.014367  -4.566 4.98e-06 ***
LowerLimitAge                 -0.044709   0.002617 -17.087  < 2e-16 ***
BLS_FEDERAL_OtherSep_Rate      3.296321   0.292822  11.257  < 2e-16 ***
BLS_FEDERAL_Quits_Rate         0.456051   0.312130   1.461  0.14399    
BLS_FEDERAL_JobOpenings_Level  0.017213   0.001847   9.320  < 2e-16 ***
LOSSqrt                       -0.777519   0.021827 -35.623  < 2e-16 ***
SEPCount_EFDATE_OCCLog        -0.032154   0.016016  -2.008  0.04469 *  
AGELVL_B1                     -0.581429   0.219502  -2.649  0.00808 ** 
AGELVL_C1                     -0.168697   0.085256  -1.979  0.04785 *  
AGELVL_E1                      0.276367   0.067215   4.112 3.93e-05 ***
AGELVL_F1                      0.493483   0.069976   7.052 1.76e-12 ***
AGELVL_G1                      0.716618   0.073835   9.706  < 2e-16 ***
AGELVL_H1                      0.849681   0.076099  11.166  < 2e-16 ***
AGELVL_I1                      0.718722   0.080987   8.875  < 2e-16 ***
LOC_011                       -0.219095   0.165455  -1.324  0.18544    
LOC_041                        0.516227   0.134188   3.847  0.00012 ***
LOC_061                        0.410572   0.084379   4.866 1.14e-06 ***
LOC_081                        0.137027   0.123131   1.113  0.26577    
LOC_131                        0.161130   0.112796   1.429  0.15315    
LOC_171                       -0.226665   0.142201  -1.594  0.11094    
LOC_181                       -0.238346   0.249000  -0.957  0.33846    
LOC_191                        0.205952   0.317785   0.648  0.51693    
LOC_201                        0.603780   0.253686   2.380  0.01731 *  
LOC_241                       -0.115951   0.080157  -1.447  0.14802    
LOC_251                        0.140571   0.179421   0.783  0.43335    
LOC_271                        0.281081   0.235430   1.194  0.23251    
LOC_291                       -0.358617   0.166495  -2.154  0.03125 *  
LOC_301                        0.671016   0.237706   2.823  0.00476 ** 
LOC_321                        0.222217   0.241610   0.920  0.35771    
LOC_341                       -0.183412   0.193286  -0.949  0.34266    
LOC_351                        0.460975   0.148224   3.110  0.00187 ** 
LOC_381                        0.478444   0.364801   1.312  0.18968    
LOC_391                       -0.296333   0.139997  -2.117  0.03428 *  
LOC_401                        0.170449   0.162973   1.046  0.29562    
LOC_421                       -0.304371   0.141143  -2.156  0.03105 *  
LOC_441                        0.326404   0.491193   0.665  0.50636    
LOC_461                        0.804818   0.251475   3.200  0.00137 ** 
LOC_471                       -0.190120   0.207431  -0.917  0.35938    
LOC_481                        0.244937   0.089323   2.742  0.00610 ** 
LOC_491                       -0.193408   0.206310  -0.937  0.34852    
LOC_501                       -0.422945   0.591177  -0.715  0.47434    
LOC_511                       -0.158691   0.075431  -2.104  0.03540 *  
LOC_531                        0.392713   0.125648   3.126  0.00177 ** 
LOC_551                       -0.328164   0.248245  -1.322  0.18619    
TOA_151                        0.071846   0.068479   1.049  0.29410    
TOA_201                        1.128002   0.192480   5.860 4.62e-09 ***
TOA_301                        0.348580   0.087595   3.979 6.91e-05 ***
TOA_351                       -0.539216   0.285987  -1.885  0.05937 .  
TOA_381                        0.333774   0.066494   5.020 5.18e-07 ***
TOA_421                        1.149154   0.441057   2.605  0.00918 ** 
TOA_441                        2.164204   1.071701   2.019  0.04344 *  
PPGROUP_111                   -0.405948   0.116626  -3.481  0.00050 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 20683  on 14919  degrees of freedom
Residual deviance: 14547  on 14867  degrees of freedom
AIC: 14653

Number of Fisher Scoring iterations: 5

                                   VIF
GSEGRD                        1.481654
LowerLimitAge                 2.249661
BLS_FEDERAL_OtherSep_Rate     1.340748
BLS_FEDERAL_Quits_Rate        1.210664
BLS_FEDERAL_JobOpenings_Level 1.088837
LOSSqrt                       1.688752
SEPCount_EFDATE_OCCLog        1.096531
AGELVL_B                      1.163278
AGELVL_C                      1.658219
AGELVL_E                      1.403404
AGELVL_F                      1.278689
AGELVL_G                      1.265681
AGELVL_H                      1.356492
AGELVL_I                      1.513072
LOC_01                        1.038549
LOC_04                        1.080522
LOC_06                        1.155708
LOC_08                        1.071440
LOC_13                        1.085967
LOC_17                        1.049719
LOC_18                        1.021546
LOC_19                        1.012762
LOC_20                        1.019301
LOC_24                        1.200105
LOC_25                        1.030274
LOC_27                        1.022506
LOC_29                        1.041619
LOC_30                        1.027878
LOC_32                        1.022898
LOC_34                        1.031483
LOC_35                        1.064436
LOC_38                        1.012088
LOC_39                        1.054960
LOC_40                        1.042594
LOC_42                        1.054934
LOC_44                        1.005838
LOC_46                        1.027085
LOC_47                        1.026188
LOC_48                        1.153324
LOC_49                        1.026005
LOC_50                        1.004394
LOC_51                        1.207361
LOC_53                        1.072411
LOC_55                        1.023498
TOA_15                        1.380722
TOA_20                        1.025931
TOA_30                        1.203909
TOA_35                        1.072500
TOA_38                        1.150034
TOA_42                        1.009406
TOA_44                        1.002972
PPGROUP_11                    1.147358

Removed AFTER this step: LOC_19 


Call:
glm(formula = fla, family = "binomial", data = OPMAnalysisDataNoFamBinary)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-2.8214  -0.7410  -0.1383   0.8191   3.1978  

Coefficients:
                               Estimate Std. Error z value Pr(>|z|)    
(Intercept)                    2.173687   0.332390   6.540 6.17e-11 ***
GSEGRD                        -0.065812   0.014364  -4.582 4.61e-06 ***
LowerLimitAge                 -0.044688   0.002616 -17.080  < 2e-16 ***
BLS_FEDERAL_OtherSep_Rate      3.295948   0.292824  11.256  < 2e-16 ***
BLS_FEDERAL_Quits_Rate         0.455946   0.312132   1.461 0.144084    
BLS_FEDERAL_JobOpenings_Level  0.017221   0.001847   9.324  < 2e-16 ***
LOSSqrt                       -0.777456   0.021825 -35.622  < 2e-16 ***
SEPCount_EFDATE_OCCLog        -0.032370   0.016013  -2.021 0.043232 *  
AGELVL_B1                     -0.580426   0.219521  -2.644 0.008192 ** 
AGELVL_C1                     -0.168460   0.085252  -1.976 0.048152 *  
AGELVL_E1                      0.276674   0.067215   4.116 3.85e-05 ***
AGELVL_F1                      0.493645   0.069977   7.054 1.73e-12 ***
AGELVL_G1                      0.716168   0.073830   9.700  < 2e-16 ***
AGELVL_H1                      0.849582   0.076095  11.165  < 2e-16 ***
AGELVL_I1                      0.718710   0.080984   8.875  < 2e-16 ***
LOC_011                       -0.222029   0.165395  -1.342 0.179461    
LOC_041                        0.513322   0.134108   3.828 0.000129 ***
LOC_061                        0.407596   0.084253   4.838 1.31e-06 ***
LOC_081                        0.134257   0.123056   1.091 0.275263    
LOC_131                        0.158342   0.112713   1.405 0.160075    
LOC_171                       -0.229485   0.142134  -1.615 0.106404    
LOC_181                       -0.241620   0.248948  -0.971 0.331764    
LOC_201                        0.600885   0.253637   2.369 0.017833 *  
LOC_241                       -0.118521   0.080059  -1.480 0.138761    
LOC_251                        0.137740   0.179366   0.768 0.442531    
LOC_271                        0.277864   0.235383   1.180 0.237809    
LOC_291                       -0.361699   0.166428  -2.173 0.029757 *  
LOC_301                        0.667897   0.237648   2.810 0.004947 ** 
LOC_321                        0.218889   0.241556   0.906 0.364851    
LOC_341                       -0.186328   0.193237  -0.964 0.334922    
LOC_351                        0.458155   0.148157   3.092 0.001986 ** 
LOC_381                        0.475284   0.364750   1.303 0.192561    
LOC_391                       -0.299316   0.139923  -2.139 0.032423 *  
LOC_401                        0.167599   0.162907   1.029 0.303573    
LOC_421                       -0.307211   0.141075  -2.178 0.029433 *  
LOC_441                        0.323275   0.491151   0.658 0.510411    
LOC_461                        0.801741   0.251422   3.189 0.001429 ** 
LOC_471                       -0.193225   0.207377  -0.932 0.351463    
LOC_481                        0.242071   0.089212   2.713 0.006659 ** 
LOC_491                       -0.196442   0.206255  -0.952 0.340882    
LOC_501                       -0.426012   0.591167  -0.721 0.471137    
LOC_511                       -0.161377   0.075319  -2.143 0.032147 *  
LOC_531                        0.389767   0.125562   3.104 0.001908 ** 
LOC_551                       -0.331540   0.248193  -1.336 0.181609    
TOA_151                        0.071387   0.068477   1.043 0.297180    
TOA_201                        1.127031   0.192477   5.855 4.76e-09 ***
TOA_301                        0.347554   0.087580   3.968 7.23e-05 ***
TOA_351                       -0.541090   0.285979  -1.892 0.058483 .  
TOA_381                        0.334941   0.066461   5.040 4.66e-07 ***
TOA_421                        1.147843   0.441061   2.602 0.009256 ** 
TOA_441                        2.161923   1.071684   2.017 0.043663 *  
PPGROUP_111                   -0.405656   0.116620  -3.478 0.000504 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 20683  on 14919  degrees of freedom
Residual deviance: 14548  on 14868  degrees of freedom
AIC: 14652

Number of Fisher Scoring iterations: 5

                                   VIF
GSEGRD                        1.480809
LowerLimitAge                 2.249496
BLS_FEDERAL_OtherSep_Rate     1.340705
BLS_FEDERAL_Quits_Rate        1.210674
BLS_FEDERAL_JobOpenings_Level 1.088762
LOSSqrt                       1.688682
SEPCount_EFDATE_OCCLog        1.096064
AGELVL_B                      1.163200
AGELVL_C                      1.658214
AGELVL_E                      1.403332
AGELVL_F                      1.278691
AGELVL_G                      1.265575
AGELVL_H                      1.356512
AGELVL_I                      1.513122
LOC_01                        1.037773
LOC_04                        1.079344
LOC_06                        1.152315
LOC_08                        1.070173
LOC_13                        1.084412
LOC_17                        1.048750
LOC_18                        1.021126
LOC_20                        1.018992
LOC_24                        1.197217
LOC_25                        1.029670
LOC_27                        1.022055
LOC_29                        1.040782
LOC_30                        1.027463
LOC_32                        1.022438
LOC_34                        1.030931
LOC_35                        1.063537
LOC_38                        1.011912
LOC_39                        1.053827
LOC_40                        1.041853
LOC_42                        1.053930
LOC_44                        1.005741
LOC_46                        1.026727
LOC_47                        1.025646
LOC_48                        1.150540
LOC_49                        1.025485
LOC_50                        1.004330
LOC_51                        1.203790
LOC_53                        1.071031
LOC_55                        1.023050
TOA_15                        1.380558
TOA_20                        1.025870
TOA_30                        1.203504
TOA_35                        1.072384
TOA_38                        1.149319
TOA_42                        1.009386
TOA_44                        1.002961
PPGROUP_11                    1.147375

Removed AFTER this step: LOC_44 


Call:
glm(formula = fla, family = "binomial", data = OPMAnalysisDataNoFamBinary)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-2.8213  -0.7408  -0.1384   0.8193   3.1969  

Coefficients:
                               Estimate Std. Error z value Pr(>|z|)    
(Intercept)                    2.178033   0.332322   6.554 5.60e-11 ***
GSEGRD                        -0.065972   0.014362  -4.594 4.36e-06 ***
LowerLimitAge                 -0.044655   0.002616 -17.071  < 2e-16 ***
BLS_FEDERAL_OtherSep_Rate      3.293772   0.292810  11.249  < 2e-16 ***
BLS_FEDERAL_Quits_Rate         0.457459   0.312119   1.466 0.142742    
BLS_FEDERAL_JobOpenings_Level  0.017199   0.001847   9.314  < 2e-16 ***
LOSSqrt                       -0.777372   0.021824 -35.620  < 2e-16 ***
SEPCount_EFDATE_OCCLog        -0.032424   0.016013  -2.025 0.042881 *  
AGELVL_B1                     -0.579635   0.219491  -2.641 0.008271 ** 
AGELVL_C1                     -0.168665   0.085250  -1.978 0.047875 *  
AGELVL_E1                      0.276445   0.067211   4.113 3.90e-05 ***
AGELVL_F1                      0.493083   0.069971   7.047 1.83e-12 ***
AGELVL_G1                      0.715405   0.073822   9.691  < 2e-16 ***
AGELVL_H1                      0.849099   0.076088  11.159  < 2e-16 ***
AGELVL_I1                      0.718276   0.080983   8.870  < 2e-16 ***
LOC_011                       -0.223879   0.165362  -1.354 0.175777    
LOC_041                        0.511302   0.134064   3.814 0.000137 ***
LOC_061                        0.405709   0.084199   4.818 1.45e-06 ***
LOC_081                        0.132402   0.123017   1.076 0.281797    
LOC_131                        0.156525   0.112672   1.389 0.164770    
LOC_171                       -0.231296   0.142100  -1.628 0.103588    
LOC_181                       -0.243457   0.248909  -0.978 0.328026    
LOC_201                        0.598891   0.253600   2.362 0.018199 *  
LOC_241                       -0.120178   0.080015  -1.502 0.133113    
LOC_251                        0.135983   0.179334   0.758 0.448292    
LOC_271                        0.275992   0.235356   1.173 0.240934    
LOC_291                       -0.363600   0.166395  -2.185 0.028877 *  
LOC_301                        0.665680   0.237608   2.802 0.005085 ** 
LOC_321                        0.216920   0.241521   0.898 0.369109    
LOC_341                       -0.188040   0.193211  -0.973 0.330436    
LOC_351                        0.456147   0.148115   3.080 0.002072 ** 
LOC_381                        0.473223   0.364705   1.298 0.194442    
LOC_391                       -0.301150   0.139887  -2.153 0.031334 *  
LOC_401                        0.165749   0.162871   1.018 0.308835    
LOC_421                       -0.308964   0.141043  -2.191 0.028483 *  
LOC_461                        0.799585   0.251380   3.181 0.001469 ** 
LOC_471                       -0.195177   0.207343  -0.941 0.346539    
LOC_481                        0.240169   0.089160   2.694 0.007067 ** 
LOC_491                       -0.198293   0.206220  -0.962 0.336269    
LOC_501                       -0.427852   0.591105  -0.724 0.469177    
LOC_511                       -0.163117   0.075271  -2.167 0.030230 *  
LOC_531                        0.387884   0.125521   3.090 0.002000 ** 
LOC_551                       -0.333439   0.248162  -1.344 0.179066    
TOA_151                        0.071191   0.068474   1.040 0.298486    
TOA_201                        1.126254   0.192464   5.852 4.86e-09 ***
TOA_301                        0.347180   0.087569   3.965 7.35e-05 ***
TOA_351                       -0.542215   0.285973  -1.896 0.057956 .  
TOA_381                        0.334720   0.066458   5.037 4.74e-07 ***
TOA_421                        1.147075   0.441055   2.601 0.009302 ** 
TOA_441                        2.160663   1.071681   2.016 0.043785 *  
PPGROUP_111                   -0.405472   0.116610  -3.477 0.000507 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 20683  on 14919  degrees of freedom
Residual deviance: 14548  on 14869  degrees of freedom
AIC: 14650

Number of Fisher Scoring iterations: 5

                                   VIF
GSEGRD                        1.480452
LowerLimitAge                 2.248635
BLS_FEDERAL_OtherSep_Rate     1.340564
BLS_FEDERAL_Quits_Rate        1.210606
BLS_FEDERAL_JobOpenings_Level 1.088423
LOSSqrt                       1.688586
SEPCount_EFDATE_OCCLog        1.096081
AGELVL_B                      1.163230
AGELVL_C                      1.658207
AGELVL_E                      1.403348
AGELVL_F                      1.278508
AGELVL_G                      1.265252
AGELVL_H                      1.356430
AGELVL_I                      1.513087
LOC_01                        1.037476
LOC_04                        1.078792
LOC_06                        1.151009
LOC_08                        1.069624
LOC_13                        1.083775
LOC_17                        1.048364
LOC_18                        1.021003
LOC_20                        1.018850
LOC_24                        1.196049
LOC_25                        1.029446
LOC_27                        1.021906
LOC_29                        1.040475
LOC_30                        1.027260
LOC_32                        1.022286
LOC_34                        1.030748
LOC_35                        1.063099
LOC_38                        1.011841
LOC_39                        1.053416
LOC_40                        1.041553
LOC_42                        1.053561
LOC_46                        1.026558
LOC_47                        1.025440
LOC_48                        1.149361
LOC_49                        1.025299
LOC_50                        1.004308
LOC_51                        1.202324
LOC_53                        1.070487
LOC_55                        1.022915
TOA_15                        1.380541
TOA_20                        1.025829
TOA_30                        1.203452
TOA_35                        1.072314
TOA_38                        1.149291
TOA_42                        1.009379
TOA_44                        1.002957
PPGROUP_11                    1.147387

Removed AFTER this step: LOC_50 


Call:
glm(formula = fla, family = "binomial", data = OPMAnalysisDataNoFamBinary)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-2.8216  -0.7414  -0.1384   0.8195   3.1974  

Coefficients:
                               Estimate Std. Error z value Pr(>|z|)    
(Intercept)                    2.178398   0.332333   6.555 5.57e-11 ***
GSEGRD                        -0.066011   0.014361  -4.596 4.30e-06 ***
LowerLimitAge                 -0.044682   0.002616 -17.083  < 2e-16 ***
BLS_FEDERAL_OtherSep_Rate      3.292869   0.292809  11.246  < 2e-16 ***
BLS_FEDERAL_Quits_Rate         0.457110   0.312100   1.465 0.143023    
BLS_FEDERAL_JobOpenings_Level  0.017207   0.001847   9.318  < 2e-16 ***
LOSSqrt                       -0.777182   0.021822 -35.614  < 2e-16 ***
SEPCount_EFDATE_OCCLog        -0.032426   0.016012  -2.025 0.042851 *  
AGELVL_B1                     -0.579526   0.219486  -2.640 0.008281 ** 
AGELVL_C1                     -0.168881   0.085245  -1.981 0.047578 *  
AGELVL_E1                      0.276819   0.067207   4.119 3.81e-05 ***
AGELVL_F1                      0.492897   0.069969   7.044 1.86e-12 ***
AGELVL_G1                      0.714823   0.073810   9.685  < 2e-16 ***
AGELVL_H1                      0.849528   0.076087  11.165  < 2e-16 ***
AGELVL_I1                      0.717806   0.080980   8.864  < 2e-16 ***
LOC_011                       -0.222196   0.165345  -1.344 0.179002    
LOC_041                        0.513048   0.134042   3.828 0.000129 ***
LOC_061                        0.407393   0.084166   4.840 1.30e-06 ***
LOC_081                        0.134063   0.122995   1.090 0.275718    
LOC_131                        0.158164   0.112648   1.404 0.160305    
LOC_171                       -0.229705   0.142081  -1.617 0.105939    
LOC_181                       -0.241574   0.248890  -0.971 0.331744    
LOC_201                        0.600682   0.253582   2.369 0.017847 *  
LOC_241                       -0.118571   0.079984  -1.482 0.138225    
LOC_251                        0.137579   0.179321   0.767 0.442949    
LOC_271                        0.277685   0.235339   1.180 0.238026    
LOC_291                       -0.361929   0.166375  -2.175 0.029601 *  
LOC_301                        0.667413   0.237600   2.809 0.004970 ** 
LOC_321                        0.218861   0.241502   0.906 0.364805    
LOC_341                       -0.186425   0.193195  -0.965 0.334565    
LOC_351                        0.457813   0.148098   3.091 0.001993 ** 
LOC_381                        0.474986   0.364692   1.302 0.192770    
LOC_391                       -0.299449   0.139866  -2.141 0.032276 *  
LOC_401                        0.167363   0.162855   1.028 0.304100    
LOC_421                       -0.307332   0.141023  -2.179 0.029309 *  
LOC_461                        0.801261   0.251367   3.188 0.001435 ** 
LOC_471                       -0.193380   0.207325  -0.933 0.350954    
LOC_481                        0.241900   0.089127   2.714 0.006646 ** 
LOC_491                       -0.196560   0.206203  -0.953 0.340469    
LOC_511                       -0.161523   0.075237  -2.147 0.031805 *  
LOC_531                        0.389511   0.125501   3.104 0.001912 ** 
LOC_551                       -0.331644   0.248141  -1.337 0.181382    
TOA_151                        0.070978   0.068466   1.037 0.299876    
TOA_201                        1.126715   0.192464   5.854 4.79e-09 ***
TOA_301                        0.348077   0.087561   3.975 7.03e-05 ***
TOA_351                       -0.541841   0.285966  -1.895 0.058122 .  
TOA_381                        0.333947   0.066446   5.026 5.01e-07 ***
TOA_421                        1.147696   0.441056   2.602 0.009264 ** 
TOA_441                        2.162176   1.071672   2.018 0.043636 *  
PPGROUP_111                   -0.406315   0.116605  -3.485 0.000493 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 20683  on 14919  degrees of freedom
Residual deviance: 14549  on 14870  degrees of freedom
AIC: 14649

Number of Fisher Scoring iterations: 5

                                   VIF
GSEGRD                        1.480438
LowerLimitAge                 2.248104
BLS_FEDERAL_OtherSep_Rate     1.340503
BLS_FEDERAL_Quits_Rate        1.210604
BLS_FEDERAL_JobOpenings_Level 1.088412
LOSSqrt                       1.688238
SEPCount_EFDATE_OCCLog        1.096052
AGELVL_B                      1.163210
AGELVL_C                      1.658174
AGELVL_E                      1.403241
AGELVL_F                      1.278470
AGELVL_G                      1.265036
AGELVL_H                      1.356328
AGELVL_I                      1.512796
LOC_01                        1.037272
LOC_04                        1.078435
LOC_06                        1.150122
LOC_08                        1.069244
LOC_13                        1.083333
LOC_17                        1.048109
LOC_18                        1.020890
LOC_20                        1.018751
LOC_24                        1.195107
LOC_25                        1.029288
LOC_27                        1.021805
LOC_29                        1.040272
LOC_30                        1.027153
LOC_32                        1.022159
LOC_34                        1.030606
LOC_35                        1.062835
LOC_38                        1.011793
LOC_39                        1.053115
LOC_40                        1.041352
LOC_42                        1.053284
LOC_46                        1.026469
LOC_47                        1.025291
LOC_48                        1.148528
LOC_49                        1.025159
LOC_51                        1.201261
LOC_53                        1.070134
LOC_55                        1.022810
TOA_15                        1.380607
TOA_20                        1.025817
TOA_30                        1.203214
TOA_35                        1.072308
TOA_38                        1.148895
TOA_42                        1.009373
TOA_44                        1.002954
PPGROUP_11                    1.147269

Removed AFTER this step: LOC_25 


Call:
glm(formula = fla, family = "binomial", data = OPMAnalysisDataNoFamBinary)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-2.8225  -0.7409  -0.1384   0.8199   3.1958  

Coefficients:
                               Estimate Std. Error z value Pr(>|z|)    
(Intercept)                    2.185356   0.332210   6.578 4.76e-11 ***
GSEGRD                        -0.066215   0.014358  -4.612 4.00e-06 ***
LowerLimitAge                 -0.044693   0.002616 -17.087  < 2e-16 ***
BLS_FEDERAL_OtherSep_Rate      3.296389   0.292766  11.259  < 2e-16 ***
BLS_FEDERAL_Quits_Rate         0.454744   0.312089   1.457 0.145089    
BLS_FEDERAL_JobOpenings_Level  0.017214   0.001846   9.323  < 2e-16 ***
LOSSqrt                       -0.777074   0.021821 -35.611  < 2e-16 ***
SEPCount_EFDATE_OCCLog        -0.032425   0.016012  -2.025 0.042868 *  
AGELVL_B1                     -0.577086   0.219416  -2.630 0.008536 ** 
AGELVL_C1                     -0.168796   0.085240  -1.980 0.047677 *  
AGELVL_E1                      0.276771   0.067205   4.118 3.82e-05 ***
AGELVL_F1                      0.493251   0.069963   7.050 1.79e-12 ***
AGELVL_G1                      0.714651   0.073806   9.683  < 2e-16 ***
AGELVL_H1                      0.849741   0.076088  11.168  < 2e-16 ***
AGELVL_I1                      0.718239   0.080984   8.869  < 2e-16 ***
LOC_011                       -0.227884   0.165181  -1.380 0.167710    
LOC_041                        0.507415   0.133842   3.791 0.000150 ***
LOC_061                        0.401730   0.083845   4.791 1.66e-06 ***
LOC_081                        0.128453   0.122781   1.046 0.295471    
LOC_131                        0.152561   0.112415   1.357 0.174744    
LOC_171                       -0.235348   0.141892  -1.659 0.097189 .  
LOC_181                       -0.247271   0.248785  -0.994 0.320263    
LOC_201                        0.595057   0.253471   2.348 0.018893 *  
LOC_241                       -0.123962   0.079679  -1.556 0.119764    
LOC_271                        0.271817   0.235219   1.156 0.247850    
LOC_291                       -0.367680   0.166211  -2.212 0.026958 *  
LOC_301                        0.661652   0.237479   2.786 0.005334 ** 
LOC_321                        0.213144   0.241394   0.883 0.377252    
LOC_341                       -0.192148   0.193057  -0.995 0.319595    
LOC_351                        0.452144   0.147914   3.057 0.002237 ** 
LOC_381                        0.469317   0.364615   1.287 0.198039    
LOC_391                       -0.305098   0.139676  -2.184 0.028938 *  
LOC_401                        0.161686   0.162687   0.994 0.320297    
LOC_421                       -0.313009   0.140831  -2.223 0.026244 *  
LOC_461                        0.795483   0.251254   3.166 0.001545 ** 
LOC_471                       -0.199060   0.207195  -0.961 0.336685    
LOC_481                        0.236310   0.088831   2.660 0.007809 ** 
LOC_491                       -0.202279   0.206072  -0.982 0.326300    
LOC_511                       -0.167083   0.074893  -2.231 0.025684 *  
LOC_531                        0.383776   0.125281   3.063 0.002189 ** 
LOC_551                       -0.337397   0.248035  -1.360 0.173742    
TOA_151                        0.070610   0.068463   1.031 0.302377    
TOA_201                        1.128703   0.192431   5.865 4.48e-09 ***
TOA_301                        0.348165   0.087557   3.976 7.00e-05 ***
TOA_351                       -0.544112   0.285939  -1.903 0.057054 .  
TOA_381                        0.334081   0.066438   5.028 4.94e-07 ***
TOA_421                        1.145523   0.441063   2.597 0.009399 ** 
TOA_441                        2.157064   1.071651   2.013 0.044131 *  
PPGROUP_111                   -0.406077   0.116607  -3.482 0.000497 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 20683  on 14919  degrees of freedom
Residual deviance: 14549  on 14871  degrees of freedom
AIC: 14647

Number of Fisher Scoring iterations: 5

                                   VIF
GSEGRD                        1.479926
LowerLimitAge                 2.248229
BLS_FEDERAL_OtherSep_Rate     1.340278
BLS_FEDERAL_Quits_Rate        1.210517
BLS_FEDERAL_JobOpenings_Level 1.088348
LOSSqrt                       1.688115
SEPCount_EFDATE_OCCLog        1.096060
AGELVL_B                      1.162989
AGELVL_C                      1.658226
AGELVL_E                      1.403223
AGELVL_F                      1.278456
AGELVL_G                      1.265052
AGELVL_H                      1.356344
AGELVL_I                      1.512637
LOC_01                        1.035198
LOC_04                        1.075230
LOC_06                        1.141334
LOC_08                        1.065492
LOC_13                        1.078816
LOC_17                        1.045329
LOC_18                        1.019989
LOC_20                        1.017908
LOC_24                        1.185992
LOC_27                        1.020732
LOC_29                        1.038175
LOC_30                        1.026131
LOC_32                        1.021193
LOC_34                        1.029083
LOC_35                        1.060207
LOC_38                        1.011383
LOC_39                        1.050218
LOC_40                        1.039225
LOC_42                        1.050413
LOC_46                        1.025551
LOC_47                        1.023994
LOC_48                        1.140914
LOC_49                        1.023830
LOC_51                        1.190253
LOC_53                        1.066363
LOC_55                        1.021883
TOA_15                        1.380547
TOA_20                        1.025636
TOA_30                        1.203224
TOA_35                        1.072257
TOA_38                        1.148892
TOA_42                        1.009336
TOA_44                        1.002915
PPGROUP_11                    1.147275

Removed AFTER this step: LOC_32 


Call:
glm(formula = fla, family = "binomial", data = OPMAnalysisDataNoFamBinary)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-2.8245  -0.7411  -0.1384   0.8200   3.1946  

Coefficients:
                               Estimate Std. Error z value Pr(>|z|)    
(Intercept)                    2.194898   0.332025   6.611 3.83e-11 ***
GSEGRD                        -0.066559   0.014353  -4.637 3.53e-06 ***
LowerLimitAge                 -0.044646   0.002615 -17.074  < 2e-16 ***
BLS_FEDERAL_OtherSep_Rate      3.294422   0.292759  11.253  < 2e-16 ***
BLS_FEDERAL_Quits_Rate         0.455059   0.312076   1.458 0.144794    
BLS_FEDERAL_JobOpenings_Level  0.017207   0.001846   9.319  < 2e-16 ***
LOSSqrt                       -0.777315   0.021818 -35.627  < 2e-16 ***
SEPCount_EFDATE_OCCLog        -0.032822   0.016006  -2.051 0.040311 *  
AGELVL_B1                     -0.579409   0.219416  -2.641 0.008274 ** 
AGELVL_C1                     -0.169572   0.085235  -1.989 0.046649 *  
AGELVL_E1                      0.275971   0.067200   4.107 4.01e-05 ***
AGELVL_F1                      0.493541   0.069952   7.055 1.72e-12 ***
AGELVL_G1                      0.714063   0.073809   9.674  < 2e-16 ***
AGELVL_H1                      0.849351   0.076082  11.164  < 2e-16 ***
AGELVL_I1                      0.717852   0.080973   8.865  < 2e-16 ***
LOC_011                       -0.232622   0.165105  -1.409 0.158855    
LOC_041                        0.502422   0.133723   3.757 0.000172 ***
LOC_061                        0.396723   0.083658   4.742 2.11e-06 ***
LOC_081                        0.123703   0.122669   1.008 0.313248    
LOC_131                        0.147944   0.112299   1.317 0.187701    
LOC_171                       -0.240024   0.141802  -1.693 0.090519 .  
LOC_181                       -0.252851   0.248721  -1.017 0.309342    
LOC_201                        0.590109   0.253422   2.329 0.019882 *  
LOC_241                       -0.128216   0.079537  -1.612 0.106955    
LOC_271                        0.266441   0.235166   1.133 0.257219    
LOC_291                       -0.372864   0.166125  -2.244 0.024802 *  
LOC_301                        0.656398   0.237414   2.765 0.005696 ** 
LOC_341                       -0.196806   0.192993  -1.020 0.307844    
LOC_351                        0.447424   0.147818   3.027 0.002471 ** 
LOC_381                        0.464045   0.364566   1.273 0.203064    
LOC_391                       -0.310056   0.139572  -2.221 0.026319 *  
LOC_401                        0.157015   0.162602   0.966 0.334222    
LOC_421                       -0.317628   0.140742  -2.257 0.024020 *  
LOC_461                        0.790372   0.251192   3.146 0.001652 ** 
LOC_471                       -0.204409   0.207127  -0.987 0.323702    
LOC_481                        0.231477   0.088665   2.611 0.009036 ** 
LOC_491                       -0.207352   0.206009  -1.007 0.314166    
LOC_511                       -0.171539   0.074731  -2.295 0.021709 *  
LOC_531                        0.379017   0.125164   3.028 0.002460 ** 
LOC_551                       -0.343258   0.247968  -1.384 0.166273    
TOA_151                        0.070483   0.068464   1.029 0.303253    
TOA_201                        1.131682   0.192515   5.878 4.14e-09 ***
TOA_301                        0.347246   0.087547   3.966 7.30e-05 ***
TOA_351                       -0.546459   0.285947  -1.911 0.055998 .  
TOA_381                        0.336618   0.066371   5.072 3.94e-07 ***
TOA_421                        1.143223   0.441064   2.592 0.009543 ** 
TOA_441                        2.153369   1.071639   2.009 0.044493 *  
PPGROUP_111                   -0.404996   0.116596  -3.473 0.000514 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 20683  on 14919  degrees of freedom
Residual deviance: 14550  on 14872  degrees of freedom
AIC: 14646

Number of Fisher Scoring iterations: 5

                                   VIF
GSEGRD                        1.478702
LowerLimitAge                 2.247078
BLS_FEDERAL_OtherSep_Rate     1.340209
BLS_FEDERAL_Quits_Rate        1.210590
BLS_FEDERAL_JobOpenings_Level 1.088248
LOSSqrt                       1.687936
SEPCount_EFDATE_OCCLog        1.095150
AGELVL_B                      1.162859
AGELVL_C                      1.657980
AGELVL_E                      1.402918
AGELVL_F                      1.278470
AGELVL_G                      1.264804
AGELVL_H                      1.356124
AGELVL_I                      1.512502
LOC_01                        1.034110
LOC_04                        1.073342
LOC_06                        1.136166
LOC_08                        1.063477
LOC_13                        1.076514
LOC_17                        1.043897
LOC_18                        1.019338
LOC_20                        1.017415
LOC_24                        1.181750
LOC_27                        1.020054
LOC_29                        1.036893
LOC_30                        1.025489
LOC_34                        1.028337
LOC_35                        1.058841
LOC_38                        1.011119
LOC_39                        1.048542
LOC_40                        1.038150
LOC_42                        1.048995
LOC_46                        1.025012
LOC_47                        1.023128
LOC_48                        1.136639
LOC_49                        1.023045
LOC_51                        1.184976
LOC_53                        1.064429
LOC_55                        1.021160
TOA_15                        1.380415
TOA_20                        1.025419
TOA_30                        1.203014
TOA_35                        1.072185
TOA_38                        1.146874
TOA_42                        1.009303
TOA_44                        1.002899
PPGROUP_11                    1.147169

Removed AFTER this step: LOC_40 


Call:
glm(formula = fla, family = "binomial", data = OPMAnalysisDataNoFamBinary)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-2.8241  -0.7412  -0.1384   0.8202   3.1926  

Coefficients:
                               Estimate Std. Error z value Pr(>|z|)    
(Intercept)                    2.212758   0.331501   6.675 2.47e-11 ***
GSEGRD                        -0.067298   0.014331  -4.696 2.65e-06 ***
LowerLimitAge                 -0.044626   0.002615 -17.068  < 2e-16 ***
BLS_FEDERAL_OtherSep_Rate      3.294398   0.292735  11.254  < 2e-16 ***
BLS_FEDERAL_Quits_Rate         0.454056   0.312076   1.455 0.145683    
BLS_FEDERAL_JobOpenings_Level  0.017205   0.001846   9.319  < 2e-16 ***
LOSSqrt                       -0.777262   0.021819 -35.623  < 2e-16 ***
SEPCount_EFDATE_OCCLog        -0.032525   0.016002  -2.033 0.042100 *  
AGELVL_B1                     -0.581892   0.219396  -2.652 0.007996 ** 
AGELVL_C1                     -0.168129   0.085219  -1.973 0.048507 *  
AGELVL_E1                      0.276947   0.067196   4.122 3.76e-05 ***
AGELVL_F1                      0.494631   0.069935   7.073 1.52e-12 ***
AGELVL_G1                      0.713681   0.073804   9.670  < 2e-16 ***
AGELVL_H1                      0.850017   0.076076  11.173  < 2e-16 ***
AGELVL_I1                      0.718792   0.080967   8.878  < 2e-16 ***
LOC_011                       -0.240166   0.164931  -1.456 0.145348    
LOC_041                        0.493798   0.133428   3.701 0.000215 ***
LOC_061                        0.389110   0.083291   4.672 2.99e-06 ***
LOC_081                        0.115838   0.122411   0.946 0.343994    
LOC_131                        0.140046   0.112006   1.250 0.211173    
LOC_171                       -0.247612   0.141587  -1.749 0.080320 .  
LOC_181                       -0.260299   0.248612  -1.047 0.295095    
LOC_201                        0.581818   0.253290   2.297 0.021617 *  
LOC_241                       -0.135643   0.079171  -1.713 0.086657 .  
LOC_271                        0.258507   0.235027   1.100 0.271375    
LOC_291                       -0.380790   0.165928  -2.295 0.021738 *  
LOC_301                        0.647695   0.237263   2.730 0.006336 ** 
LOC_341                       -0.204434   0.192840  -1.060 0.289091    
LOC_351                        0.438787   0.147560   2.974 0.002943 ** 
LOC_381                        0.455605   0.364497   1.250 0.211316    
LOC_391                       -0.317642   0.139356  -2.279 0.022645 *  
LOC_421                       -0.325120   0.140531  -2.314 0.020694 *  
LOC_461                        0.781369   0.251032   3.113 0.001854 ** 
LOC_471                       -0.212136   0.206966  -1.025 0.305373    
LOC_481                        0.223435   0.088279   2.531 0.011374 *  
LOC_491                       -0.215332   0.205853  -1.046 0.295539    
LOC_511                       -0.179331   0.074307  -2.413 0.015806 *  
LOC_531                        0.370934   0.124890   2.970 0.002977 ** 
LOC_551                       -0.350540   0.247847  -1.414 0.157262    
TOA_151                        0.070652   0.068462   1.032 0.302079    
TOA_201                        1.129217   0.192541   5.865 4.50e-09 ***
TOA_301                        0.345442   0.087499   3.948 7.88e-05 ***
TOA_351                       -0.548742   0.285874  -1.920 0.054918 .  
TOA_381                        0.334593   0.066335   5.044 4.56e-07 ***
TOA_421                        1.140319   0.441075   2.585 0.009729 ** 
TOA_441                        2.147433   1.071610   2.004 0.045077 *  
PPGROUP_111                   -0.408296   0.116506  -3.504 0.000457 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 20683  on 14919  degrees of freedom
Residual deviance: 14551  on 14873  degrees of freedom
AIC: 14645

Number of Fisher Scoring iterations: 5

                                   VIF
GSEGRD                        1.474094
LowerLimitAge                 2.246894
BLS_FEDERAL_OtherSep_Rate     1.340107
BLS_FEDERAL_Quits_Rate        1.210536
BLS_FEDERAL_JobOpenings_Level 1.088166
LOSSqrt                       1.688079
SEPCount_EFDATE_OCCLog        1.094754
AGELVL_B                      1.162661
AGELVL_C                      1.657351
AGELVL_E                      1.402507
AGELVL_F                      1.278284
AGELVL_G                      1.264822
AGELVL_H                      1.356156
AGELVL_I                      1.512463
LOC_01                        1.031813
LOC_04                        1.068566
LOC_06                        1.126183
LOC_08                        1.058826
LOC_13                        1.070850
LOC_17                        1.040736
LOC_18                        1.018376
LOC_20                        1.016253
LOC_24                        1.170786
LOC_27                        1.018829
LOC_29                        1.034399
LOC_30                        1.024005
LOC_34                        1.026644
LOC_35                        1.054978
LOC_38                        1.010541
LOC_39                        1.045264
LOC_42                        1.045865
LOC_46                        1.023603
LOC_47                        1.021623
LOC_48                        1.126688
LOC_49                        1.021417
LOC_51                        1.171317
LOC_53                        1.059714
LOC_55                        1.020241
TOA_15                        1.380371
TOA_20                        1.025235
TOA_30                        1.202260
TOA_35                        1.072083
TOA_38                        1.145634
TOA_42                        1.009265
TOA_44                        1.002866
PPGROUP_11                    1.146138

Removed AFTER this step: LOC_08 


Call:
glm(formula = fla, family = "binomial", data = OPMAnalysisDataNoFamBinary)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-2.8251  -0.7410  -0.1385   0.8202   3.1898  

Coefficients:
                               Estimate Std. Error z value Pr(>|z|)    
(Intercept)                    2.226907   0.331142   6.725 1.76e-11 ***
GSEGRD                        -0.067798   0.014322  -4.734 2.20e-06 ***
LowerLimitAge                 -0.044612   0.002615 -17.062  < 2e-16 ***
BLS_FEDERAL_OtherSep_Rate      3.292260   0.292708  11.248  < 2e-16 ***
BLS_FEDERAL_Quits_Rate         0.457848   0.312029   1.467 0.142289    
BLS_FEDERAL_JobOpenings_Level  0.017193   0.001846   9.313  < 2e-16 ***
LOSSqrt                       -0.777026   0.021815 -35.619  < 2e-16 ***
SEPCount_EFDATE_OCCLog        -0.032260   0.015999  -2.016 0.043756 *  
AGELVL_B1                     -0.585577   0.219366  -2.669 0.007599 ** 
AGELVL_C1                     -0.169029   0.085216  -1.984 0.047308 *  
AGELVL_E1                      0.276792   0.067193   4.119 3.80e-05 ***
AGELVL_F1                      0.494216   0.069926   7.068 1.58e-12 ***
AGELVL_G1                      0.714636   0.073806   9.683  < 2e-16 ***
AGELVL_H1                      0.850740   0.076053  11.186  < 2e-16 ***
AGELVL_I1                      0.718288   0.080973   8.871  < 2e-16 ***
LOC_011                       -0.249365   0.164638  -1.515 0.129867    
LOC_041                        0.483733   0.133002   3.637 0.000276 ***
LOC_061                        0.379766   0.082703   4.592 4.39e-06 ***
LOC_131                        0.130493   0.111551   1.170 0.242078    
LOC_171                       -0.256802   0.141244  -1.818 0.069043 .  
LOC_181                       -0.269230   0.248412  -1.084 0.278452    
LOC_201                        0.571855   0.253077   2.260 0.023845 *  
LOC_241                       -0.144733   0.078588  -1.842 0.065524 .  
LOC_271                        0.248972   0.234803   1.060 0.288988    
LOC_291                       -0.390336   0.165618  -2.357 0.018431 *  
LOC_301                        0.637463   0.237019   2.690 0.007156 ** 
LOC_341                       -0.213335   0.192601  -1.108 0.268013    
LOC_351                        0.428721   0.147175   2.913 0.003580 ** 
LOC_381                        0.445638   0.364337   1.223 0.221273    
LOC_391                       -0.326822   0.139010  -2.351 0.018720 *  
LOC_421                       -0.334140   0.140201  -2.383 0.017158 *  
LOC_461                        0.770976   0.250781   3.074 0.002110 ** 
LOC_471                       -0.221455   0.206720  -1.071 0.284043    
LOC_481                        0.213721   0.087681   2.437 0.014789 *  
LOC_491                       -0.224866   0.205604  -1.094 0.274093    
LOC_511                       -0.188720   0.073650  -2.562 0.010396 *  
LOC_531                        0.361285   0.124473   2.903 0.003702 ** 
LOC_551                       -0.359539   0.247646  -1.452 0.146550    
TOA_151                        0.071768   0.068454   1.048 0.294451    
TOA_201                        1.131611   0.192559   5.877 4.19e-09 ***
TOA_301                        0.342589   0.087457   3.917 8.96e-05 ***
TOA_351                       -0.552669   0.285874  -1.933 0.053204 .  
TOA_381                        0.332616   0.066297   5.017 5.25e-07 ***
TOA_421                        1.136968   0.441078   2.578 0.009946 ** 
TOA_441                        2.139424   1.071581   1.997 0.045878 *  
PPGROUP_111                   -0.409204   0.116506  -3.512 0.000444 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 20683  on 14919  degrees of freedom
Residual deviance: 14552  on 14874  degrees of freedom
AIC: 14644

Number of Fisher Scoring iterations: 5

                                   VIF
GSEGRD                        1.471910
LowerLimitAge                 2.246864
BLS_FEDERAL_OtherSep_Rate     1.339985
BLS_FEDERAL_Quits_Rate        1.210308
BLS_FEDERAL_JobOpenings_Level 1.088069
LOSSqrt                       1.687659
SEPCount_EFDATE_OCCLog        1.094374
AGELVL_B                      1.162369
AGELVL_C                      1.657200
AGELVL_E                      1.402616
AGELVL_F                      1.278362
AGELVL_G                      1.264575
AGELVL_H                      1.356430
AGELVL_I                      1.512358
LOC_01                        1.028245
LOC_04                        1.061789
LOC_06                        1.110438
LOC_13                        1.062187
LOC_17                        1.035886
LOC_18                        1.016929
LOC_20                        1.014500
LOC_24                        1.153692
LOC_27                        1.016974
LOC_29                        1.030603
LOC_30                        1.021874
LOC_34                        1.024240
LOC_35                        1.049512
LOC_38                        1.009706
LOC_39                        1.040243
LOC_42                        1.041116
LOC_46                        1.021645
LOC_47                        1.019330
LOC_48                        1.111513
LOC_49                        1.018985
LOC_51                        1.150611
LOC_53                        1.052689
LOC_55                        1.018762
TOA_15                        1.379999
TOA_20                        1.025042
TOA_30                        1.200666
TOA_35                        1.071887
TOA_38                        1.144507
TOA_42                        1.009210
TOA_44                        1.002803
PPGROUP_11                    1.146013

Removed AFTER this step: TOA_15 


Call:
glm(formula = fla, family = "binomial", data = OPMAnalysisDataNoFamBinary)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-2.8323  -0.7406  -0.1371   0.8196   3.1978  

Coefficients:
                               Estimate Std. Error z value Pr(>|z|)    
(Intercept)                    2.284875   0.326509   6.998 2.60e-12 ***
GSEGRD                        -0.069789   0.014189  -4.918 8.72e-07 ***
LowerLimitAge                 -0.044400   0.002606 -17.038  < 2e-16 ***
BLS_FEDERAL_OtherSep_Rate      3.293957   0.292777  11.251  < 2e-16 ***
BLS_FEDERAL_Quits_Rate         0.457266   0.311974   1.466 0.142725    
BLS_FEDERAL_JobOpenings_Level  0.017202   0.001846   9.321  < 2e-16 ***
LOSSqrt                       -0.786305   0.019980 -39.355  < 2e-16 ***
SEPCount_EFDATE_OCCLog        -0.031685   0.015983  -1.982 0.047433 *  
AGELVL_B1                     -0.565726   0.218381  -2.591 0.009582 ** 
AGELVL_C1                     -0.168417   0.085158  -1.978 0.047963 *  
AGELVL_E1                      0.278146   0.067148   4.142 3.44e-05 ***
AGELVL_F1                      0.496050   0.069879   7.099 1.26e-12 ***
AGELVL_G1                      0.716975   0.073778   9.718  < 2e-16 ***
AGELVL_H1                      0.852558   0.076067  11.208  < 2e-16 ***
AGELVL_I1                      0.721587   0.080969   8.912  < 2e-16 ***
LOC_011                       -0.250835   0.164729  -1.523 0.127830    
LOC_041                        0.484749   0.132912   3.647 0.000265 ***
LOC_061                        0.378318   0.082724   4.573 4.80e-06 ***
LOC_131                        0.129046   0.111516   1.157 0.247190    
LOC_171                       -0.259971   0.141227  -1.841 0.065651 .  
LOC_181                       -0.266334   0.248365  -1.072 0.283562    
LOC_201                        0.576614   0.252964   2.279 0.022642 *  
LOC_241                       -0.143268   0.078551  -1.824 0.068170 .  
LOC_271                        0.248636   0.234774   1.059 0.289580    
LOC_291                       -0.393320   0.165620  -2.375 0.017557 *  
LOC_301                        0.635658   0.237057   2.681 0.007330 ** 
LOC_341                       -0.214443   0.192700  -1.113 0.265779    
LOC_351                        0.429215   0.147113   2.918 0.003528 ** 
LOC_381                        0.444926   0.364179   1.222 0.221813    
LOC_391                       -0.326725   0.139036  -2.350 0.018777 *  
LOC_421                       -0.335981   0.140213  -2.396 0.016565 *  
LOC_461                        0.769164   0.250718   3.068 0.002156 ** 
LOC_471                       -0.223466   0.206729  -1.081 0.279714    
LOC_481                        0.212147   0.087672   2.420 0.015530 *  
LOC_491                       -0.223720   0.205395  -1.089 0.276056    
LOC_511                       -0.185844   0.073526  -2.528 0.011485 *  
LOC_531                        0.360595   0.124424   2.898 0.003754 ** 
LOC_551                       -0.362465   0.247744  -1.463 0.143450    
TOA_201                        1.111302   0.191863   5.792 6.95e-09 ***
TOA_301                        0.333212   0.087144   3.824 0.000131 ***
TOA_351                       -0.590290   0.283744  -2.080 0.037492 *  
TOA_381                        0.315580   0.064374   4.902 9.47e-07 ***
TOA_421                        1.116051   0.440658   2.533 0.011319 *  
TOA_441                        2.133573   1.071743   1.991 0.046508 *  
PPGROUP_111                   -0.418532   0.116104  -3.605 0.000312 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 20683  on 14919  degrees of freedom
Residual deviance: 14553  on 14875  degrees of freedom
AIC: 14643

Number of Fisher Scoring iterations: 5

                                   VIF
GSEGRD                        1.446463
LowerLimitAge                 2.231811
BLS_FEDERAL_OtherSep_Rate     1.339932
BLS_FEDERAL_Quits_Rate        1.210258
BLS_FEDERAL_JobOpenings_Level 1.088234
LOSSqrt                       1.412967
SEPCount_EFDATE_OCCLog        1.093163
AGELVL_B                      1.153319
AGELVL_C                      1.657331
AGELVL_E                      1.401872
AGELVL_F                      1.277556
AGELVL_G                      1.263549
AGELVL_H                      1.355802
AGELVL_I                      1.509972
LOC_01                        1.028137
LOC_04                        1.061851
LOC_06                        1.110070
LOC_13                        1.062088
LOC_17                        1.035430
LOC_18                        1.016804
LOC_20                        1.014193
LOC_24                        1.153416
LOC_27                        1.016966
LOC_29                        1.030312
LOC_30                        1.021783
LOC_34                        1.024156
LOC_35                        1.049494
LOC_38                        1.009721
LOC_39                        1.040222
LOC_42                        1.040943
LOC_46                        1.021593
LOC_47                        1.019240
LOC_48                        1.111186
LOC_49                        1.019008
LOC_51                        1.149432
LOC_53                        1.052728
LOC_55                        1.018643
TOA_20                        1.014651
TOA_30                        1.187990
TOA_35                        1.054678
TOA_38                        1.075975
TOA_42                        1.007166
TOA_44                        1.002774
PPGROUP_11                    1.139765

Removed AFTER this step: LOC_27 


Call:
glm(formula = fla, family = "binomial", data = OPMAnalysisDataNoFamBinary)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-2.8336  -0.7410  -0.1372   0.8201   3.1970  

Coefficients:
                               Estimate Std. Error z value Pr(>|z|)    
(Intercept)                    2.300668   0.326157   7.054 1.74e-12 ***
GSEGRD                        -0.070581   0.014170  -4.981 6.32e-07 ***
LowerLimitAge                 -0.044447   0.002606 -17.059  < 2e-16 ***
BLS_FEDERAL_OtherSep_Rate      3.295665   0.292769  11.257  < 2e-16 ***
BLS_FEDERAL_Quits_Rate         0.456128   0.311950   1.462 0.143692    
BLS_FEDERAL_JobOpenings_Level  0.017221   0.001845   9.333  < 2e-16 ***
LOSSqrt                       -0.785966   0.019975 -39.348  < 2e-16 ***
SEPCount_EFDATE_OCCLog        -0.031742   0.015980  -1.986 0.046994 *  
AGELVL_B1                     -0.565670   0.218332  -2.591 0.009573 ** 
AGELVL_C1                     -0.169989   0.085146  -1.996 0.045886 *  
AGELVL_E1                      0.277912   0.067145   4.139 3.49e-05 ***
AGELVL_F1                      0.496279   0.069870   7.103 1.22e-12 ***
AGELVL_G1                      0.717127   0.073770   9.721  < 2e-16 ***
AGELVL_H1                      0.852763   0.076065  11.211  < 2e-16 ***
AGELVL_I1                      0.721527   0.080966   8.911  < 2e-16 ***
LOC_011                       -0.256069   0.164678  -1.555 0.119953    
LOC_041                        0.479225   0.132815   3.608 0.000308 ***
LOC_061                        0.373008   0.082582   4.517 6.28e-06 ***
LOC_131                        0.123943   0.111423   1.112 0.265980    
LOC_171                       -0.265129   0.141155  -1.878 0.060342 .  
LOC_181                       -0.272070   0.248331  -1.096 0.273256    
LOC_201                        0.571208   0.252935   2.258 0.023926 *  
LOC_241                       -0.147705   0.078447  -1.883 0.059720 .  
LOC_291                       -0.399017   0.165550  -2.410 0.015941 *  
LOC_301                        0.629863   0.237013   2.658 0.007872 ** 
LOC_341                       -0.219640   0.192653  -1.140 0.254251    
LOC_351                        0.423717   0.147036   2.882 0.003955 ** 
LOC_381                        0.439068   0.364170   1.206 0.227945    
LOC_391                       -0.332207   0.138955  -2.391 0.016814 *  
LOC_421                       -0.341250   0.140134  -2.435 0.014885 *  
LOC_461                        0.763145   0.250674   3.044 0.002332 ** 
LOC_471                       -0.229106   0.206682  -1.108 0.267648    
LOC_481                        0.206884   0.087539   2.363 0.018111 *  
LOC_491                       -0.229129   0.205349  -1.116 0.264507    
LOC_511                       -0.190784   0.073391  -2.600 0.009334 ** 
LOC_531                        0.355131   0.124325   2.856 0.004284 ** 
LOC_551                       -0.368640   0.247701  -1.488 0.136686    
TOA_201                        1.111609   0.191918   5.792 6.95e-09 ***
TOA_301                        0.332203   0.087146   3.812 0.000138 ***
TOA_351                       -0.594681   0.283732  -2.096 0.036089 *  
TOA_381                        0.317449   0.064335   4.934 8.04e-07 ***
TOA_421                        1.114271   0.440688   2.528 0.011456 *  
TOA_441                        2.130680   1.071735   1.988 0.046804 *  
PPGROUP_111                   -0.420026   0.116101  -3.618 0.000297 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 20683  on 14919  degrees of freedom
Residual deviance: 14554  on 14876  degrees of freedom
AIC: 14642

Number of Fisher Scoring iterations: 5

                                   VIF
GSEGRD                        1.442709
LowerLimitAge                 2.231148
BLS_FEDERAL_OtherSep_Rate     1.339987
BLS_FEDERAL_Quits_Rate        1.210271
BLS_FEDERAL_JobOpenings_Level 1.088134
LOSSqrt                       1.412203
SEPCount_EFDATE_OCCLog        1.093173
AGELVL_B                      1.153381
AGELVL_C                      1.656663
AGELVL_E                      1.401770
AGELVL_F                      1.277569
AGELVL_G                      1.263577
AGELVL_H                      1.355800
AGELVL_I                      1.509930
LOC_01                        1.027220
LOC_04                        1.060244
LOC_06                        1.106053
LOC_13                        1.060145
LOC_17                        1.034230
LOC_18                        1.016329
LOC_20                        1.013785
LOC_24                        1.150238
LOC_29                        1.029246
LOC_30                        1.021240
LOC_34                        1.023515
LOC_35                        1.048198
LOC_38                        1.009493
LOC_39                        1.038804
LOC_42                        1.039667
LOC_46                        1.021069
LOC_47                        1.018572
LOC_48                        1.107684
LOC_49                        1.018392
LOC_51                        1.144940
LOC_53                        1.050951
LOC_55                        1.018087
TOA_20                        1.014639
TOA_30                        1.187852
TOA_35                        1.054407
TOA_38                        1.075292
TOA_42                        1.007154
TOA_44                        1.002767
PPGROUP_11                    1.139572

Removed AFTER this step: LOC_18 


Call:
glm(formula = fla, family = "binomial", data = OPMAnalysisDataNoFamBinary)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-2.8336  -0.7405  -0.1369   0.8203   3.1984  

Coefficients:
                               Estimate Std. Error z value Pr(>|z|)    
(Intercept)                    2.301357   0.326167   7.056 1.72e-12 ***
GSEGRD                        -0.070413   0.014169  -4.970 6.71e-07 ***
LowerLimitAge                 -0.044479   0.002605 -17.072  < 2e-16 ***
BLS_FEDERAL_OtherSep_Rate      3.296003   0.292769  11.258  < 2e-16 ***
BLS_FEDERAL_Quits_Rate         0.449135   0.311856   1.440 0.149811    
BLS_FEDERAL_JobOpenings_Level  0.017208   0.001845   9.326  < 2e-16 ***
LOSSqrt                       -0.786147   0.019972 -39.362  < 2e-16 ***
SEPCount_EFDATE_OCCLog        -0.031777   0.015980  -1.988 0.046757 *  
AGELVL_B1                     -0.564125   0.218307  -2.584 0.009764 ** 
AGELVL_C1                     -0.171217   0.085135  -2.011 0.044313 *  
AGELVL_E1                      0.278207   0.067141   4.144 3.42e-05 ***
AGELVL_F1                      0.496232   0.069867   7.102 1.23e-12 ***
AGELVL_G1                      0.719013   0.073749   9.749  < 2e-16 ***
AGELVL_H1                      0.854416   0.076043  11.236  < 2e-16 ***
AGELVL_I1                      0.722358   0.080969   8.921  < 2e-16 ***
LOC_011                       -0.251154   0.164606  -1.526 0.127061    
LOC_041                        0.483948   0.132751   3.646 0.000267 ***
LOC_061                        0.377941   0.082455   4.584 4.57e-06 ***
LOC_131                        0.128419   0.111346   1.153 0.248773    
LOC_171                       -0.260465   0.141080  -1.846 0.064859 .  
LOC_201                        0.575914   0.252898   2.277 0.022771 *  
LOC_241                       -0.143501   0.078351  -1.832 0.067022 .  
LOC_291                       -0.393978   0.165474  -2.381 0.017271 *  
LOC_301                        0.634601   0.236984   2.678 0.007410 ** 
LOC_341                       -0.214724   0.192593  -1.115 0.264888    
LOC_351                        0.428411   0.146977   2.915 0.003559 ** 
LOC_381                        0.443876   0.364125   1.219 0.222837    
LOC_391                       -0.327121   0.138869  -2.356 0.018493 *  
LOC_421                       -0.336346   0.140058  -2.401 0.016329 *  
LOC_461                        0.768163   0.250645   3.065 0.002179 ** 
LOC_471                       -0.223872   0.206609  -1.084 0.278562    
LOC_481                        0.211590   0.087432   2.420 0.015519 *  
LOC_491                       -0.224250   0.205287  -1.092 0.274669    
LOC_511                       -0.186489   0.073281  -2.545 0.010933 *  
LOC_531                        0.359848   0.124253   2.896 0.003778 ** 
LOC_551                       -0.362699   0.247606  -1.465 0.142970    
TOA_201                        1.111397   0.191913   5.791 6.99e-09 ***
TOA_301                        0.333743   0.087136   3.830 0.000128 ***
TOA_351                       -0.592970   0.283704  -2.090 0.036609 *  
TOA_381                        0.313658   0.064233   4.883 1.04e-06 ***
TOA_421                        1.115476   0.440672   2.531 0.011364 *  
TOA_441                        2.134395   1.071723   1.992 0.046420 *  
PPGROUP_111                   -0.421494   0.116111  -3.630 0.000283 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 20683  on 14919  degrees of freedom
Residual deviance: 14555  on 14877  degrees of freedom
AIC: 14641

Number of Fisher Scoring iterations: 5

                                   VIF
GSEGRD                        1.442606
LowerLimitAge                 2.231047
BLS_FEDERAL_OtherSep_Rate     1.339971
BLS_FEDERAL_Quits_Rate        1.209863
BLS_FEDERAL_JobOpenings_Level 1.088031
LOSSqrt                       1.412203
SEPCount_EFDATE_OCCLog        1.093142
AGELVL_B                      1.153281
AGELVL_C                      1.656357
AGELVL_E                      1.401727
AGELVL_F                      1.277552
AGELVL_G                      1.262980
AGELVL_H                      1.355447
AGELVL_I                      1.509768
LOC_01                        1.026454
LOC_04                        1.059092
LOC_06                        1.102722
LOC_13                        1.058690
LOC_17                        1.033273
LOC_20                        1.013485
LOC_24                        1.147435
LOC_29                        1.028436
LOC_30                        1.020893
LOC_34                        1.022944
LOC_35                        1.047286
LOC_38                        1.009342
LOC_39                        1.037628
LOC_42                        1.038581
LOC_46                        1.020717
LOC_47                        1.018020
LOC_48                        1.104957
LOC_49                        1.017906
LOC_51                        1.141582
LOC_53                        1.049660
LOC_55                        1.017595
TOA_20                        1.014600
TOA_30                        1.187619
TOA_35                        1.054347
TOA_38                        1.072010
TOA_42                        1.007143
TOA_44                        1.002758
PPGROUP_11                    1.139384

Removed AFTER this step: LOC_47 


Call:
glm(formula = fla, family = "binomial", data = OPMAnalysisDataNoFamBinary)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-2.8334  -0.7413  -0.1372   0.8200   3.2002  

Coefficients:
                               Estimate Std. Error z value Pr(>|z|)    
(Intercept)                    2.288947   0.325910   7.023 2.17e-12 ***
GSEGRD                        -0.069972   0.014162  -4.941 7.78e-07 ***
LowerLimitAge                 -0.044529   0.002605 -17.092  < 2e-16 ***
BLS_FEDERAL_OtherSep_Rate      3.299628   0.292740  11.272  < 2e-16 ***
BLS_FEDERAL_Quits_Rate         0.446453   0.311816   1.432 0.152205    
BLS_FEDERAL_JobOpenings_Level  0.017222   0.001845   9.335  < 2e-16 ***
LOSSqrt                       -0.786202   0.019970 -39.369  < 2e-16 ***
SEPCount_EFDATE_OCCLog        -0.031781   0.015979  -1.989 0.046714 *  
AGELVL_B1                     -0.564885   0.218235  -2.588 0.009642 ** 
AGELVL_C1                     -0.170353   0.085129  -2.001 0.045379 *  
AGELVL_E1                      0.278912   0.067136   4.154 3.26e-05 ***
AGELVL_F1                      0.496473   0.069861   7.107 1.19e-12 ***
AGELVL_G1                      0.720110   0.073744   9.765  < 2e-16 ***
AGELVL_H1                      0.856021   0.076022  11.260  < 2e-16 ***
AGELVL_I1                      0.723021   0.080974   8.929  < 2e-16 ***
LOC_011                       -0.245622   0.164524  -1.493 0.135455    
LOC_041                        0.490164   0.132634   3.696 0.000219 ***
LOC_061                        0.383615   0.082290   4.662 3.14e-06 ***
LOC_131                        0.134002   0.111229   1.205 0.228304    
LOC_171                       -0.255039   0.140990  -1.809 0.070465 .  
LOC_201                        0.581832   0.252840   2.301 0.021381 *  
LOC_241                       -0.138352   0.078206  -1.769 0.076882 .  
LOC_291                       -0.388107   0.165381  -2.347 0.018939 *  
LOC_301                        0.640959   0.236913   2.705 0.006821 ** 
LOC_341                       -0.209339   0.192531  -1.087 0.276905    
LOC_351                        0.434482   0.146874   2.958 0.003094 ** 
LOC_381                        0.450007   0.364087   1.236 0.216463    
LOC_391                       -0.321330   0.138766  -2.316 0.020579 *  
LOC_421                       -0.330852   0.139968  -2.364 0.018090 *  
LOC_461                        0.774567   0.250587   3.091 0.001995 ** 
LOC_481                        0.217435   0.087267   2.492 0.012717 *  
LOC_491                       -0.218615   0.205227  -1.065 0.286769    
LOC_511                       -0.181206   0.073113  -2.478 0.013196 *  
LOC_531                        0.365528   0.124149   2.944 0.003237 ** 
LOC_551                       -0.356485   0.247535  -1.440 0.149828    
TOA_201                        1.111930   0.191968   5.792 6.94e-09 ***
TOA_301                        0.334832   0.087134   3.843 0.000122 ***
TOA_351                       -0.589765   0.283667  -2.079 0.037611 *  
TOA_381                        0.312202   0.064217   4.862 1.16e-06 ***
TOA_421                        1.117389   0.440662   2.536 0.011222 *  
TOA_441                        2.138344   1.071717   1.995 0.046016 *  
PPGROUP_111                   -0.419447   0.116082  -3.613 0.000302 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 20683  on 14919  degrees of freedom
Residual deviance: 14556  on 14878  degrees of freedom
AIC: 14640

Number of Fisher Scoring iterations: 5

                                   VIF
GSEGRD                        1.441478
LowerLimitAge                 2.230723
BLS_FEDERAL_OtherSep_Rate     1.339728
BLS_FEDERAL_Quits_Rate        1.209663
BLS_FEDERAL_JobOpenings_Level 1.088000
LOSSqrt                       1.412143
SEPCount_EFDATE_OCCLog        1.093141
AGELVL_B                      1.153341
AGELVL_C                      1.656215
AGELVL_E                      1.401687
AGELVL_F                      1.277689
AGELVL_G                      1.262870
AGELVL_H                      1.355321
AGELVL_I                      1.510002
LOC_01                        1.025462
LOC_04                        1.057098
LOC_06                        1.098240
LOC_13                        1.056395
LOC_17                        1.031951
LOC_20                        1.013009
LOC_24                        1.143145
LOC_29                        1.027318
LOC_30                        1.020271
LOC_34                        1.022244
LOC_35                        1.045751
LOC_38                        1.009095
LOC_39                        1.036075
LOC_42                        1.037193
LOC_46                        1.020145
LOC_48                        1.100711
LOC_49                        1.017242
LOC_51                        1.136432
LOC_53                        1.047766
LOC_55                        1.017044
TOA_20                        1.014611
TOA_30                        1.187469
TOA_35                        1.054241
TOA_38                        1.071490
TOA_42                        1.007120
TOA_44                        1.002746
PPGROUP_11                    1.139105

Removed AFTER this step: LOC_49 


Call:
glm(formula = fla, family = "binomial", data = OPMAnalysisDataNoFamBinary)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-2.8326  -0.7416  -0.1372   0.8194   3.2019  

Coefficients:
                               Estimate Std. Error z value Pr(>|z|)    
(Intercept)                    2.279301   0.325773   6.997 2.62e-12 ***
GSEGRD                        -0.069673   0.014159  -4.921 8.62e-07 ***
LowerLimitAge                 -0.044545   0.002605 -17.101  < 2e-16 ***
BLS_FEDERAL_OtherSep_Rate      3.299455   0.292722  11.272  < 2e-16 ***
BLS_FEDERAL_Quits_Rate         0.451147   0.311784   1.447 0.147900    
BLS_FEDERAL_JobOpenings_Level  0.017219   0.001845   9.333  < 2e-16 ***
LOSSqrt                       -0.786288   0.019970 -39.374  < 2e-16 ***
SEPCount_EFDATE_OCCLog        -0.031662   0.015978  -1.982 0.047529 *  
AGELVL_B1                     -0.563039   0.218217  -2.580 0.009875 ** 
AGELVL_C1                     -0.171589   0.085118  -2.016 0.043810 *  
AGELVL_E1                      0.278193   0.067126   4.144 3.41e-05 ***
AGELVL_F1                      0.496255   0.069865   7.103 1.22e-12 ***
AGELVL_G1                      0.720114   0.073742   9.765  < 2e-16 ***
AGELVL_H1                      0.855562   0.076011  11.256  < 2e-16 ***
AGELVL_I1                      0.721648   0.080954   8.914  < 2e-16 ***
LOC_011                       -0.240162   0.164448  -1.460 0.144177    
LOC_041                        0.495796   0.132532   3.741 0.000183 ***
LOC_061                        0.389091   0.082130   4.737 2.16e-06 ***
LOC_131                        0.139458   0.111112   1.255 0.209439    
LOC_171                       -0.249710   0.140907  -1.772 0.076368 .  
LOC_201                        0.587493   0.252799   2.324 0.020128 *  
LOC_241                       -0.133149   0.078053  -1.706 0.088030 .  
LOC_291                       -0.382416   0.165295  -2.314 0.020693 *  
LOC_301                        0.646869   0.236860   2.731 0.006314 ** 
LOC_341                       -0.203926   0.192461  -1.060 0.289340    
LOC_351                        0.440065   0.146783   2.998 0.002717 ** 
LOC_381                        0.455832   0.364070   1.252 0.210554    
LOC_391                       -0.315969   0.138678  -2.278 0.022700 *  
LOC_421                       -0.325593   0.139882  -2.328 0.019932 *  
LOC_461                        0.780279   0.250534   3.114 0.001843 ** 
LOC_481                        0.222916   0.087116   2.559 0.010502 *  
LOC_511                       -0.175804   0.072933  -2.410 0.015931 *  
LOC_531                        0.371172   0.124039   2.992 0.002768 ** 
LOC_551                       -0.350992   0.247486  -1.418 0.156124    
TOA_201                        1.110839   0.192020   5.785 7.25e-09 ***
TOA_301                        0.337484   0.087101   3.875 0.000107 ***
TOA_351                       -0.587042   0.283639  -2.070 0.038482 *  
TOA_381                        0.312530   0.064216   4.867 1.13e-06 ***
TOA_421                        1.119737   0.440646   2.541 0.011050 *  
TOA_441                        2.143004   1.071728   2.000 0.045546 *  
PPGROUP_111                   -0.419963   0.116068  -3.618 0.000297 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 20683  on 14919  degrees of freedom
Residual deviance: 14557  on 14879  degrees of freedom
AIC: 14639

Number of Fisher Scoring iterations: 5

                                   VIF
GSEGRD                        1.440964
LowerLimitAge                 2.230539
BLS_FEDERAL_OtherSep_Rate     1.339622
BLS_FEDERAL_Quits_Rate        1.209313
BLS_FEDERAL_JobOpenings_Level 1.088000
LOSSqrt                       1.412197
SEPCount_EFDATE_OCCLog        1.093097
AGELVL_B                      1.153190
AGELVL_C                      1.655688
AGELVL_E                      1.401531
AGELVL_F                      1.277535
AGELVL_G                      1.262762
AGELVL_H                      1.355273
AGELVL_I                      1.509508
LOC_01                        1.024460
LOC_04                        1.055396
LOC_06                        1.093927
LOC_13                        1.054135
LOC_17                        1.030631
LOC_20                        1.012557
LOC_24                        1.138629
LOC_29                        1.026235
LOC_30                        1.019709
LOC_34                        1.021518
LOC_35                        1.044403
LOC_38                        1.008862
LOC_39                        1.034693
LOC_42                        1.035878
LOC_46                        1.019673
LOC_48                        1.096847
LOC_51                        1.130880
LOC_53                        1.045840
LOC_55                        1.016591
TOA_20                        1.014529
TOA_30                        1.186588
TOA_35                        1.054118
TOA_38                        1.071478
TOA_42                        1.007089
TOA_44                        1.002729
PPGROUP_11                    1.139079

Removed AFTER this step: LOC_34 


Call:
glm(formula = fla, family = "binomial", data = OPMAnalysisDataNoFamBinary)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-2.8326  -0.7421  -0.1372   0.8190   3.2036  

Coefficients:
                               Estimate Std. Error z value Pr(>|z|)    
(Intercept)                    2.273147   0.325680   6.980 2.96e-12 ***
GSEGRD                        -0.069754   0.014159  -4.927 8.37e-07 ***
LowerLimitAge                 -0.044516   0.002605 -17.091  < 2e-16 ***
BLS_FEDERAL_OtherSep_Rate      3.297426   0.292713  11.265  < 2e-16 ***
BLS_FEDERAL_Quits_Rate         0.451452   0.311778   1.448 0.147618    
BLS_FEDERAL_JobOpenings_Level  0.017222   0.001845   9.336  < 2e-16 ***
LOSSqrt                       -0.786517   0.019967 -39.390  < 2e-16 ***
SEPCount_EFDATE_OCCLog        -0.031516   0.015978  -1.973 0.048551 *  
AGELVL_B1                     -0.569066   0.218030  -2.610 0.009053 ** 
AGELVL_C1                     -0.174880   0.085038  -2.056 0.039736 *  
AGELVL_E1                      0.278584   0.067129   4.150 3.33e-05 ***
AGELVL_F1                      0.495746   0.069856   7.097 1.28e-12 ***
AGELVL_G1                      0.720380   0.073743   9.769  < 2e-16 ***
AGELVL_H1                      0.855502   0.076011  11.255  < 2e-16 ***
AGELVL_I1                      0.722020   0.080959   8.918  < 2e-16 ***
LOC_011                       -0.234490   0.164350  -1.427 0.153645    
LOC_041                        0.500919   0.132441   3.782 0.000155 ***
LOC_061                        0.394540   0.081963   4.814 1.48e-06 ***
LOC_131                        0.144818   0.110989   1.305 0.191963    
LOC_171                       -0.244189   0.140798  -1.734 0.082862 .  
LOC_201                        0.592712   0.252739   2.345 0.019019 *  
LOC_241                       -0.127418   0.077857  -1.637 0.101719    
LOC_291                       -0.377062   0.165209  -2.282 0.022469 *  
LOC_301                        0.651917   0.236812   2.753 0.005907 ** 
LOC_351                        0.445382   0.146684   3.036 0.002395 ** 
LOC_381                        0.461173   0.364045   1.267 0.205226    
LOC_391                       -0.310309   0.138559  -2.240 0.025120 *  
LOC_421                       -0.319727   0.139758  -2.288 0.022154 *  
LOC_461                        0.785319   0.250490   3.135 0.001718 ** 
LOC_481                        0.228151   0.086971   2.623 0.008708 ** 
LOC_511                       -0.170127   0.072727  -2.339 0.019321 *  
LOC_531                        0.376910   0.123906   3.042 0.002351 ** 
LOC_551                       -0.345354   0.247416  -1.396 0.162761    
TOA_201                        1.110535   0.191952   5.785 7.23e-09 ***
TOA_301                        0.339884   0.087059   3.904 9.46e-05 ***
TOA_351                       -0.593078   0.283491  -2.092 0.036434 *  
TOA_381                        0.312282   0.064213   4.863 1.15e-06 ***
TOA_421                        1.121734   0.440621   2.546 0.010903 *  
TOA_441                        2.148620   1.071715   2.005 0.044980 *  
PPGROUP_111                   -0.418722   0.116028  -3.609 0.000308 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 20683  on 14919  degrees of freedom
Residual deviance: 14559  on 14880  degrees of freedom
AIC: 14639

Number of Fisher Scoring iterations: 5

                                   VIF
GSEGRD                        1.441092
LowerLimitAge                 2.230622
BLS_FEDERAL_OtherSep_Rate     1.339593
BLS_FEDERAL_Quits_Rate        1.209410
BLS_FEDERAL_JobOpenings_Level 1.087977
LOSSqrt                       1.412039
SEPCount_EFDATE_OCCLog        1.093007
AGELVL_B                      1.152407
AGELVL_C                      1.653729
AGELVL_E                      1.401295
AGELVL_F                      1.277467
AGELVL_G                      1.262676
AGELVL_H                      1.355241
AGELVL_I                      1.509331
LOC_01                        1.023371
LOC_04                        1.053965
LOC_06                        1.089602
LOC_13                        1.051921
LOC_17                        1.029202
LOC_20                        1.012167
LOC_24                        1.133096
LOC_29                        1.025262
LOC_30                        1.019297
LOC_35                        1.043170
LOC_38                        1.008663
LOC_39                        1.033141
LOC_42                        1.034230
LOC_46                        1.019298
LOC_48                        1.093275
LOC_51                        1.124673
LOC_53                        1.043836
LOC_55                        1.016109
TOA_20                        1.014544
TOA_30                        1.185853
TOA_35                        1.053524
TOA_38                        1.071444
TOA_42                        1.007066
TOA_44                        1.002704
PPGROUP_11                    1.138933

Removed AFTER this step: LOC_38 


Call:
glm(formula = fla, family = "binomial", data = OPMAnalysisDataNoFamBinary)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-2.8331  -0.7421  -0.1371   0.8188   3.2035  

Coefficients:
                               Estimate Std. Error z value Pr(>|z|)    
(Intercept)                    2.288737   0.325439   7.033 2.02e-12 ***
GSEGRD                        -0.070649   0.014141  -4.996 5.85e-07 ***
LowerLimitAge                 -0.044472   0.002604 -17.078  < 2e-16 ***
BLS_FEDERAL_OtherSep_Rate      3.297244   0.292701  11.265  < 2e-16 ***
BLS_FEDERAL_Quits_Rate         0.457003   0.311725   1.466 0.142635    
BLS_FEDERAL_JobOpenings_Level  0.017221   0.001845   9.335  < 2e-16 ***
LOSSqrt                       -0.786469   0.019962 -39.397  < 2e-16 ***
SEPCount_EFDATE_OCCLog        -0.031999   0.015973  -2.003 0.045144 *  
AGELVL_B1                     -0.571840   0.218034  -2.623 0.008723 ** 
AGELVL_C1                     -0.174323   0.085040  -2.050 0.040375 *  
AGELVL_E1                      0.279111   0.067125   4.158 3.21e-05 ***
AGELVL_F1                      0.495977   0.069848   7.101 1.24e-12 ***
AGELVL_G1                      0.721810   0.073732   9.790  < 2e-16 ***
AGELVL_H1                      0.855262   0.076008  11.252  < 2e-16 ***
AGELVL_I1                      0.722810   0.080945   8.930  < 2e-16 ***
LOC_011                       -0.238067   0.164338  -1.449 0.147438    
LOC_041                        0.496477   0.132394   3.750 0.000177 ***
LOC_061                        0.390781   0.081910   4.771 1.83e-06 ***
LOC_131                        0.141148   0.110952   1.272 0.203317    
LOC_171                       -0.247711   0.140773  -1.760 0.078468 .  
LOC_201                        0.588347   0.252720   2.328 0.019909 *  
LOC_241                       -0.130644   0.077820  -1.679 0.093194 .  
LOC_291                       -0.380884   0.165185  -2.306 0.021122 *  
LOC_301                        0.646994   0.236773   2.733 0.006285 ** 
LOC_351                        0.441029   0.146651   3.007 0.002636 ** 
LOC_391                       -0.314029   0.138532  -2.267 0.023401 *  
LOC_421                       -0.323189   0.139737  -2.313 0.020732 *  
LOC_461                        0.780437   0.250460   3.116 0.001833 ** 
LOC_481                        0.224234   0.086918   2.580 0.009885 ** 
LOC_511                       -0.173680   0.072684  -2.390 0.016870 *  
LOC_531                        0.372853   0.123865   3.010 0.002611 ** 
LOC_551                       -0.349191   0.247385  -1.412 0.158089    
TOA_201                        1.108204   0.191942   5.774 7.76e-09 ***
TOA_301                        0.340300   0.087034   3.910 9.23e-05 ***
TOA_351                       -0.596965   0.283500  -2.106 0.035230 *  
TOA_381                        0.311270   0.064192   4.849 1.24e-06 ***
TOA_421                        1.137031   0.440440   2.582 0.009835 ** 
TOA_441                        2.148009   1.071708   2.004 0.045039 *  
PPGROUP_111                   -0.422552   0.115970  -3.644 0.000269 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 20683  on 14919  degrees of freedom
Residual deviance: 14560  on 14881  degrees of freedom
AIC: 14638

Number of Fisher Scoring iterations: 5

                                   VIF
GSEGRD                        1.437401
LowerLimitAge                 2.229757
BLS_FEDERAL_OtherSep_Rate     1.339492
BLS_FEDERAL_Quits_Rate        1.209139
BLS_FEDERAL_JobOpenings_Level 1.087965
LOSSqrt                       1.411618
SEPCount_EFDATE_OCCLog        1.092337
AGELVL_B                      1.152332
AGELVL_C                      1.653555
AGELVL_E                      1.401297
AGELVL_F                      1.277537
AGELVL_G                      1.262409
AGELVL_H                      1.355159
AGELVL_I                      1.509413
LOC_01                        1.023075
LOC_04                        1.053224
LOC_06                        1.088228
LOC_13                        1.051226
LOC_17                        1.028819
LOC_20                        1.011976
LOC_24                        1.131941
LOC_29                        1.024934
LOC_30                        1.019018
LOC_35                        1.042590
LOC_39                        1.032692
LOC_42                        1.033845
LOC_46                        1.019052
LOC_48                        1.091917
LOC_51                        1.123080
LOC_53                        1.043162
LOC_55                        1.015969
TOA_20                        1.014444
TOA_30                        1.185656
TOA_35                        1.053449
TOA_38                        1.071210
TOA_42                        1.006519
TOA_44                        1.002704
PPGROUP_11                    1.138163

Removed AFTER this step: LOC_13 


Call:
glm(formula = fla, family = "binomial", data = OPMAnalysisDataNoFamBinary)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-2.8328  -0.7414  -0.1372   0.8197   3.1996  

Coefficients:
                               Estimate Std. Error z value Pr(>|z|)    
(Intercept)                    2.305021   0.325167   7.089 1.35e-12 ***
GSEGRD                        -0.070803   0.014141  -5.007 5.53e-07 ***
LowerLimitAge                 -0.044453   0.002604 -17.071  < 2e-16 ***
BLS_FEDERAL_OtherSep_Rate      3.295540   0.292640  11.261  < 2e-16 ***
BLS_FEDERAL_Quits_Rate         0.454130   0.311663   1.457 0.145083    
BLS_FEDERAL_JobOpenings_Level  0.017240   0.001845   9.346  < 2e-16 ***
LOSSqrt                       -0.786405   0.019959 -39.401  < 2e-16 ***
SEPCount_EFDATE_OCCLog        -0.031412   0.015965  -1.968 0.049118 *  
AGELVL_B1                     -0.571853   0.217993  -2.623 0.008709 ** 
AGELVL_C1                     -0.175126   0.085037  -2.059 0.039455 *  
AGELVL_E1                      0.279596   0.067122   4.165 3.11e-05 ***
AGELVL_F1                      0.496321   0.069849   7.106 1.20e-12 ***
AGELVL_G1                      0.722116   0.073718   9.796  < 2e-16 ***
AGELVL_H1                      0.855454   0.075991  11.257  < 2e-16 ***
AGELVL_I1                      0.723392   0.080948   8.937  < 2e-16 ***
LOC_011                       -0.249110   0.164089  -1.518 0.128980    
LOC_041                        0.484148   0.132031   3.667 0.000245 ***
LOC_061                        0.379716   0.081443   4.662 3.13e-06 ***
LOC_171                       -0.258659   0.140490  -1.841 0.065604 .  
LOC_201                        0.576610   0.252545   2.283 0.022419 *  
LOC_241                       -0.142290   0.077279  -1.841 0.065586 .  
LOC_291                       -0.392200   0.164931  -2.378 0.017408 *  
LOC_301                        0.634859   0.236584   2.683 0.007287 ** 
LOC_351                        0.428815   0.146332   2.930 0.003385 ** 
LOC_391                       -0.325077   0.138246  -2.351 0.018701 *  
LOC_421                       -0.334075   0.139463  -2.395 0.016601 *  
LOC_461                        0.768194   0.250260   3.070 0.002144 ** 
LOC_481                        0.212320   0.086411   2.457 0.014006 *  
LOC_511                       -0.185081   0.072134  -2.566 0.010294 *  
LOC_531                        0.361283   0.123526   2.925 0.003447 ** 
LOC_551                       -0.359414   0.247226  -1.454 0.146006    
TOA_201                        1.108519   0.191835   5.779 7.54e-09 ***
TOA_301                        0.334690   0.086931   3.850 0.000118 ***
TOA_351                       -0.600851   0.283472  -2.120 0.034039 *  
TOA_381                        0.308129   0.064127   4.805 1.55e-06 ***
TOA_421                        1.133616   0.440402   2.574 0.010052 *  
TOA_441                        2.136459   1.071660   1.994 0.046196 *  
PPGROUP_111                   -0.428159   0.115916  -3.694 0.000221 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 20683  on 14919  degrees of freedom
Residual deviance: 14562  on 14882  degrees of freedom
AIC: 14638

Number of Fisher Scoring iterations: 5

                                   VIF
GSEGRD                        1.437018
LowerLimitAge                 2.229701
BLS_FEDERAL_OtherSep_Rate     1.339322
BLS_FEDERAL_Quits_Rate        1.209083
BLS_FEDERAL_JobOpenings_Level 1.087880
LOSSqrt                       1.411623
SEPCount_EFDATE_OCCLog        1.091343
AGELVL_B                      1.152194
AGELVL_C                      1.653433
AGELVL_E                      1.401295
AGELVL_F                      1.277524
AGELVL_G                      1.262512
AGELVL_H                      1.355362
AGELVL_I                      1.509405
LOC_01                        1.020234
LOC_04                        1.047587
LOC_06                        1.076039
LOC_17                        1.025023
LOC_20                        1.010633
LOC_24                        1.116448
LOC_29                        1.021994
LOC_30                        1.017351
LOC_35                        1.038133
LOC_39                        1.028675
LOC_42                        1.030027
LOC_46                        1.017548
LOC_48                        1.079290
LOC_51                        1.106252
LOC_53                        1.037582
LOC_55                        1.014920
TOA_20                        1.014397
TOA_30                        1.182600
TOA_35                        1.053292
TOA_38                        1.069624
TOA_42                        1.006494
TOA_44                        1.002632
PPGROUP_11                    1.136280

Removed AFTER this step: LOC_55 


Call:
glm(formula = fla, family = "binomial", data = OPMAnalysisDataNoFamBinary)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-2.8304  -0.7424  -0.1372   0.8191   3.2006  

Coefficients:
                               Estimate Std. Error z value Pr(>|z|)    
(Intercept)                    2.290859   0.324962   7.050 1.79e-12 ***
GSEGRD                        -0.070260   0.014133  -4.972 6.64e-07 ***
LowerLimitAge                 -0.044406   0.002604 -17.056  < 2e-16 ***
BLS_FEDERAL_OtherSep_Rate      3.304905   0.292561  11.296  < 2e-16 ***
BLS_FEDERAL_Quits_Rate         0.447237   0.311599   1.435 0.151202    
BLS_FEDERAL_JobOpenings_Level  0.017242   0.001845   9.348  < 2e-16 ***
LOSSqrt                       -0.786744   0.019961 -39.414  < 2e-16 ***
SEPCount_EFDATE_OCCLog        -0.031099   0.015960  -1.949 0.051348 .  
AGELVL_B1                     -0.569395   0.217909  -2.613 0.008975 ** 
AGELVL_C1                     -0.173741   0.085016  -2.044 0.040989 *  
AGELVL_E1                      0.281498   0.067106   4.195 2.73e-05 ***
AGELVL_F1                      0.498374   0.069823   7.138 9.49e-13 ***
AGELVL_G1                      0.723541   0.073704   9.817  < 2e-16 ***
AGELVL_H1                      0.856734   0.075979  11.276  < 2e-16 ***
AGELVL_I1                      0.724646   0.080951   8.952  < 2e-16 ***
LOC_011                       -0.243686   0.164018  -1.486 0.137351    
LOC_041                        0.489189   0.131983   3.706 0.000210 ***
LOC_061                        0.385230   0.081346   4.736 2.18e-06 ***
LOC_171                       -0.253347   0.140420  -1.804 0.071199 .  
LOC_201                        0.581768   0.252498   2.304 0.021220 *  
LOC_241                       -0.138029   0.077217  -1.788 0.073849 .  
LOC_291                       -0.386277   0.164852  -2.343 0.019120 *  
LOC_301                        0.640153   0.236546   2.706 0.006805 ** 
LOC_351                        0.433813   0.146281   2.966 0.003021 ** 
LOC_391                       -0.319236   0.138168  -2.310 0.020861 *  
LOC_421                       -0.328519   0.139396  -2.357 0.018436 *  
LOC_461                        0.773537   0.250218   3.091 0.001992 ** 
LOC_481                        0.217318   0.086337   2.517 0.011833 *  
LOC_511                       -0.180347   0.072050  -2.503 0.012311 *  
LOC_531                        0.366559   0.123469   2.969 0.002989 ** 
TOA_201                        1.106144   0.191853   5.766 8.14e-09 ***
TOA_301                        0.335294   0.086920   3.857 0.000115 ***
TOA_351                       -0.597167   0.283439  -2.107 0.035130 *  
TOA_381                        0.302901   0.064015   4.732 2.23e-06 ***
TOA_421                        1.133028   0.440466   2.572 0.010101 *  
TOA_441                        2.139104   1.071653   1.996 0.045925 *  
PPGROUP_111                   -0.429486   0.115917  -3.705 0.000211 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 20683  on 14919  degrees of freedom
Residual deviance: 14564  on 14883  degrees of freedom
AIC: 14638

Number of Fisher Scoring iterations: 5

                                   VIF
GSEGRD                        1.435977
LowerLimitAge                 2.229347
BLS_FEDERAL_OtherSep_Rate     1.338907
BLS_FEDERAL_Quits_Rate        1.208788
BLS_FEDERAL_JobOpenings_Level 1.087945
LOSSqrt                       1.411848
SEPCount_EFDATE_OCCLog        1.091253
AGELVL_B                      1.152168
AGELVL_C                      1.653359
AGELVL_E                      1.400710
AGELVL_F                      1.277035
AGELVL_G                      1.262333
AGELVL_H                      1.355247
AGELVL_I                      1.509236
LOC_01                        1.019713
LOC_04                        1.046827
LOC_06                        1.073693
LOC_17                        1.024316
LOC_20                        1.010429
LOC_24                        1.114786
LOC_29                        1.021359
LOC_30                        1.017101
LOC_35                        1.037548
LOC_39                        1.027798
LOC_42                        1.029238
LOC_46                        1.017318
LOC_48                        1.077540
LOC_51                        1.103901
LOC_53                        1.036661
TOA_20                        1.014267
TOA_30                        1.182683
TOA_35                        1.053172
TOA_38                        1.065754
TOA_42                        1.006473
TOA_44                        1.002628
PPGROUP_11                    1.136089

Removed AFTER this step: BLS_FEDERAL_Quits_Rate 


Call:
glm(formula = fla, family = "binomial", data = OPMAnalysisDataNoFamBinary)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-2.8339  -0.7430  -0.1386   0.8205   3.1825  

Coefficients:
                               Estimate Std. Error z value Pr(>|z|)    
(Intercept)                    2.410808   0.314127   7.675 1.66e-14 ***
GSEGRD                        -0.070410   0.014130  -4.983 6.26e-07 ***
LowerLimitAge                 -0.044487   0.002603 -17.092  < 2e-16 ***
BLS_FEDERAL_OtherSep_Rate      3.476209   0.267379  13.001  < 2e-16 ***
BLS_FEDERAL_JobOpenings_Level  0.017455   0.001840   9.487  < 2e-16 ***
LOSSqrt                       -0.787113   0.019958 -39.439  < 2e-16 ***
SEPCount_EFDATE_OCCLog        -0.030764   0.015957  -1.928 0.053873 .  
AGELVL_B1                     -0.564717   0.217824  -2.593 0.009527 ** 
AGELVL_C1                     -0.174318   0.085000  -2.051 0.040287 *  
AGELVL_E1                      0.281212   0.067103   4.191 2.78e-05 ***
AGELVL_F1                      0.497914   0.069818   7.132 9.92e-13 ***
AGELVL_G1                      0.724310   0.073687   9.829  < 2e-16 ***
AGELVL_H1                      0.856605   0.075975  11.275  < 2e-16 ***
AGELVL_I1                      0.724080   0.080943   8.946  < 2e-16 ***
LOC_011                       -0.244474   0.164008  -1.491 0.136061    
LOC_041                        0.492762   0.131928   3.735 0.000188 ***
LOC_061                        0.385996   0.081350   4.745 2.09e-06 ***
LOC_171                       -0.252538   0.140354  -1.799 0.071972 .  
LOC_201                        0.586532   0.252542   2.323 0.020205 *  
LOC_241                       -0.137812   0.077207  -1.785 0.074267 .  
LOC_291                       -0.387618   0.164849  -2.351 0.018705 *  
LOC_301                        0.639158   0.236775   2.699 0.006946 ** 
LOC_351                        0.435321   0.146296   2.976 0.002924 ** 
LOC_391                       -0.317610   0.138174  -2.299 0.021526 *  
LOC_421                       -0.328389   0.139440  -2.355 0.018520 *  
LOC_461                        0.777644   0.250480   3.105 0.001905 ** 
LOC_481                        0.218995   0.086324   2.537 0.011184 *  
LOC_511                       -0.181523   0.072039  -2.520 0.011742 *  
LOC_531                        0.368145   0.123466   2.982 0.002866 ** 
TOA_201                        1.104861   0.191762   5.762 8.33e-09 ***
TOA_301                        0.336799   0.086915   3.875 0.000107 ***
TOA_351                       -0.595588   0.283317  -2.102 0.035536 *  
TOA_381                        0.305170   0.064005   4.768 1.86e-06 ***
TOA_421                        1.130290   0.440326   2.567 0.010260 *  
TOA_441                        2.136749   1.071254   1.995 0.046084 *  
PPGROUP_111                   -0.429273   0.115908  -3.704 0.000213 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 20683  on 14919  degrees of freedom
Residual deviance: 14566  on 14884  degrees of freedom
AIC: 14638

Number of Fisher Scoring iterations: 5

                                   VIF
GSEGRD                        1.435633
LowerLimitAge                 2.228674
BLS_FEDERAL_OtherSep_Rate     1.117995
BLS_FEDERAL_JobOpenings_Level 1.081106
LOSSqrt                       1.411994
SEPCount_EFDATE_OCCLog        1.090958
AGELVL_B                      1.151897
AGELVL_C                      1.653567
AGELVL_E                      1.400770
AGELVL_F                      1.277023
AGELVL_G                      1.262300
AGELVL_H                      1.354998
AGELVL_I                      1.508727
LOC_01                        1.019717
LOC_04                        1.046466
LOC_06                        1.073594
LOC_17                        1.024308
LOC_20                        1.010226
LOC_24                        1.114790
LOC_29                        1.021318
LOC_30                        1.017084
LOC_35                        1.037466
LOC_39                        1.027712
LOC_42                        1.029216
LOC_46                        1.017137
LOC_48                        1.077366
LOC_51                        1.103727
LOC_53                        1.036590
TOA_20                        1.014292
TOA_30                        1.182450
TOA_35                        1.053153
TOA_38                        1.065029
TOA_42                        1.006464
TOA_44                        1.002625
PPGROUP_11                    1.136123

Removed AFTER this step: LOC_01 


Call:
glm(formula = fla, family = "binomial", data = OPMAnalysisDataNoFamBinary)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-2.8357  -0.7426  -0.1396   0.8210   3.1864  

Coefficients:
                               Estimate Std. Error z value Pr(>|z|)    
(Intercept)                    2.404722   0.314060   7.657 1.90e-14 ***
GSEGRD                        -0.070372   0.014131  -4.980 6.36e-07 ***
LowerLimitAge                 -0.044455   0.002602 -17.082  < 2e-16 ***
BLS_FEDERAL_OtherSep_Rate      3.473632   0.267317  12.994  < 2e-16 ***
BLS_FEDERAL_JobOpenings_Level  0.017480   0.001840   9.502  < 2e-16 ***
LOSSqrt                       -0.788198   0.019948 -39.513  < 2e-16 ***
SEPCount_EFDATE_OCCLog        -0.030964   0.015958  -1.940 0.052331 .  
AGELVL_B1                     -0.566431   0.217699  -2.602 0.009271 ** 
AGELVL_C1                     -0.175700   0.084973  -2.068 0.038667 *  
AGELVL_E1                      0.282576   0.067091   4.212 2.53e-05 ***
AGELVL_F1                      0.499843   0.069814   7.160 8.09e-13 ***
AGELVL_G1                      0.724842   0.073691   9.836  < 2e-16 ***
AGELVL_H1                      0.856697   0.075970  11.277  < 2e-16 ***
AGELVL_I1                      0.723699   0.080935   8.942  < 2e-16 ***
LOC_041                        0.500660   0.131854   3.797 0.000146 ***
LOC_061                        0.394184   0.081180   4.856 1.20e-06 ***
LOC_171                       -0.244415   0.140275  -1.742 0.081439 .  
LOC_201                        0.594707   0.252563   2.355 0.018538 *  
LOC_241                       -0.129963   0.077038  -1.687 0.091603 .  
LOC_291                       -0.379545   0.164794  -2.303 0.021270 *  
LOC_301                        0.647414   0.236797   2.734 0.006256 ** 
LOC_351                        0.443536   0.146225   3.033 0.002419 ** 
LOC_391                       -0.309232   0.138084  -2.239 0.025127 *  
LOC_421                       -0.320052   0.139352  -2.297 0.021634 *  
LOC_461                        0.785760   0.250486   3.137 0.001707 ** 
LOC_481                        0.227017   0.086174   2.634 0.008429 ** 
LOC_511                       -0.173676   0.071855  -2.417 0.015647 *  
LOC_531                        0.376442   0.123363   3.052 0.002277 ** 
TOA_201                        1.107347   0.191649   5.778 7.56e-09 ***
TOA_301                        0.339791   0.086900   3.910 9.22e-05 ***
TOA_351                       -0.592705   0.283301  -2.092 0.036427 *  
TOA_381                        0.304139   0.063998   4.752 2.01e-06 ***
TOA_421                        1.127037   0.440279   2.560 0.010473 *  
TOA_441                        2.143992   1.071237   2.001 0.045347 *  
PPGROUP_111                   -0.430148   0.115913  -3.711 0.000206 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 20683  on 14919  degrees of freedom
Residual deviance: 14568  on 14885  degrees of freedom
AIC: 14638

Number of Fisher Scoring iterations: 5

                                   VIF
GSEGRD                        1.435790
LowerLimitAge                 2.228846
BLS_FEDERAL_OtherSep_Rate     1.118045
BLS_FEDERAL_JobOpenings_Level 1.080979
LOSSqrt                       1.411164
SEPCount_EFDATE_OCCLog        1.091001
AGELVL_B                      1.152035
AGELVL_C                      1.653633
AGELVL_E                      1.400446
AGELVL_F                      1.276537
AGELVL_G                      1.262359
AGELVL_H                      1.355196
AGELVL_I                      1.509271
LOC_04                        1.044765
LOC_06                        1.068700
LOC_17                        1.022750
LOC_20                        1.009747
LOC_24                        1.109539
LOC_29                        1.020207
LOC_30                        1.016531
LOC_35                        1.035993
LOC_39                        1.026006
LOC_42                        1.027542
LOC_46                        1.016655
LOC_48                        1.073172
LOC_51                        1.097751
LOC_53                        1.034479
TOA_20                        1.014194
TOA_30                        1.181889
TOA_35                        1.053121
TOA_38                        1.064879
TOA_42                        1.006467
TOA_44                        1.002604
PPGROUP_11                    1.135982

Removed AFTER this step: LOC_24 


Call:
glm(formula = fla, family = "binomial", data = OPMAnalysisDataNoFamBinary)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-2.8398  -0.7438  -0.1399   0.8201   3.1933  

Coefficients:
                               Estimate Std. Error z value Pr(>|z|)    
(Intercept)                    2.397151   0.314022   7.634 2.28e-14 ***
GSEGRD                        -0.073329   0.014025  -5.229 1.71e-07 ***
LowerLimitAge                 -0.044388   0.002602 -17.062  < 2e-16 ***
BLS_FEDERAL_OtherSep_Rate      3.478798   0.267266  13.016  < 2e-16 ***
BLS_FEDERAL_JobOpenings_Level  0.017506   0.001839   9.518  < 2e-16 ***
LOSSqrt                       -0.786901   0.019927 -39.490  < 2e-16 ***
SEPCount_EFDATE_OCCLog        -0.030636   0.015954  -1.920 0.054827 .  
AGELVL_B1                     -0.568014   0.217795  -2.608 0.009107 ** 
AGELVL_C1                     -0.177206   0.084965  -2.086 0.037012 *  
AGELVL_E1                      0.283270   0.067092   4.222 2.42e-05 ***
AGELVL_F1                      0.499108   0.069817   7.149 8.75e-13 ***
AGELVL_G1                      0.724958   0.073668   9.841  < 2e-16 ***
AGELVL_H1                      0.856695   0.075940  11.281  < 2e-16 ***
AGELVL_I1                      0.722119   0.080897   8.926  < 2e-16 ***
LOC_041                        0.518626   0.131412   3.947 7.93e-05 ***
LOC_061                        0.411642   0.080519   5.112 3.18e-07 ***
LOC_171                       -0.226480   0.139892  -1.619 0.105455    
LOC_201                        0.612054   0.252344   2.425 0.015288 *  
LOC_291                       -0.363176   0.164536  -2.207 0.027295 *  
LOC_301                        0.663652   0.236583   2.805 0.005029 ** 
LOC_351                        0.461066   0.145844   3.161 0.001570 ** 
LOC_391                       -0.291498   0.137688  -2.117 0.034252 *  
LOC_421                       -0.301898   0.138929  -2.173 0.029777 *  
LOC_461                        0.800962   0.250341   3.199 0.001377 ** 
LOC_481                        0.245487   0.085467   2.872 0.004075 ** 
LOC_511                       -0.153810   0.070889  -2.170 0.030027 *  
LOC_531                        0.394527   0.122875   3.211 0.001324 ** 
TOA_201                        1.111242   0.191572   5.801 6.61e-09 ***
TOA_301                        0.354148   0.086505   4.094 4.24e-05 ***
TOA_351                       -0.606255   0.282982  -2.142 0.032163 *  
TOA_381                        0.312748   0.063780   4.904 9.41e-07 ***
TOA_421                        1.124635   0.440332   2.554 0.010647 *  
TOA_441                        2.172956   1.071076   2.029 0.042483 *  
PPGROUP_111                   -0.418824   0.115586  -3.623 0.000291 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 20683  on 14919  degrees of freedom
Residual deviance: 14571  on 14886  degrees of freedom
AIC: 14639

Number of Fisher Scoring iterations: 5

                                   VIF
GSEGRD                        1.414132
LowerLimitAge                 2.228218
BLS_FEDERAL_OtherSep_Rate     1.118058
BLS_FEDERAL_JobOpenings_Level 1.081055
LOSSqrt                       1.408028
SEPCount_EFDATE_OCCLog        1.090862
AGELVL_B                      1.151972
AGELVL_C                      1.653603
AGELVL_E                      1.400482
AGELVL_F                      1.276403
AGELVL_G                      1.262428
AGELVL_H                      1.355066
AGELVL_I                      1.508723
LOC_04                        1.037962
LOC_06                        1.051265
LOC_17                        1.016828
LOC_20                        1.008066
LOC_29                        1.016635
LOC_30                        1.014882
LOC_35                        1.030761
LOC_39                        1.020027
LOC_42                        1.021352
LOC_46                        1.015328
LOC_48                        1.055825
LOC_51                        1.067927
LOC_53                        1.026646
TOA_20                        1.014077
TOA_30                        1.170882
TOA_35                        1.052610
TOA_38                        1.057914
TOA_42                        1.006489
TOA_44                        1.002348
PPGROUP_11                    1.132533

Removed AFTER this step: LOC_17 


Call:
glm(formula = fla, family = "binomial", data = OPMAnalysisDataNoFamBinary)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-2.8395  -0.7428  -0.1398   0.8209   3.1960  

Coefficients:
                               Estimate Std. Error z value Pr(>|z|)    
(Intercept)                    2.381260   0.313801   7.588 3.24e-14 ***
GSEGRD                        -0.072869   0.014020  -5.197 2.02e-07 ***
LowerLimitAge                 -0.044319   0.002601 -17.041  < 2e-16 ***
BLS_FEDERAL_OtherSep_Rate      3.480776   0.267208  13.026  < 2e-16 ***
BLS_FEDERAL_JobOpenings_Level  0.017533   0.001839   9.534  < 2e-16 ***
LOSSqrt                       -0.787354   0.019925 -39.516  < 2e-16 ***
SEPCount_EFDATE_OCCLog        -0.030891   0.015951  -1.937 0.052784 .  
AGELVL_B1                     -0.566213   0.217694  -2.601 0.009296 ** 
AGELVL_C1                     -0.176287   0.084947  -2.075 0.037963 *  
AGELVL_E1                      0.284571   0.067075   4.243 2.21e-05 ***
AGELVL_F1                      0.501685   0.069796   7.188 6.58e-13 ***
AGELVL_G1                      0.726202   0.073662   9.859  < 2e-16 ***
AGELVL_H1                      0.856103   0.075935  11.274  < 2e-16 ***
AGELVL_I1                      0.721935   0.080887   8.925  < 2e-16 ***
LOC_041                        0.526966   0.131316   4.013 6.00e-05 ***
LOC_061                        0.420088   0.080349   5.228 1.71e-07 ***
LOC_201                        0.620633   0.252298   2.460 0.013897 *  
LOC_291                       -0.354516   0.164448  -2.156 0.031100 *  
LOC_301                        0.672284   0.236533   2.842 0.004480 ** 
LOC_351                        0.469709   0.145746   3.223 0.001270 ** 
LOC_391                       -0.283005   0.137589  -2.057 0.039697 *  
LOC_421                       -0.293330   0.138830  -2.113 0.034612 *  
LOC_461                        0.809726   0.250283   3.235 0.001215 ** 
LOC_481                        0.253756   0.085316   2.974 0.002937 ** 
LOC_511                       -0.145413   0.070693  -2.057 0.039689 *  
LOC_531                        0.403080   0.122762   3.283 0.001025 ** 
TOA_201                        1.112386   0.191527   5.808 6.32e-09 ***
TOA_301                        0.354352   0.086517   4.096 4.21e-05 ***
TOA_351                       -0.607520   0.283087  -2.146 0.031869 *  
TOA_381                        0.312899   0.063797   4.905 9.36e-07 ***
TOA_421                        1.126043   0.440364   2.557 0.010556 *  
TOA_441                        2.179502   1.071052   2.035 0.041859 *  
PPGROUP_111                   -0.420732   0.115605  -3.639 0.000273 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 20683  on 14919  degrees of freedom
Residual deviance: 14574  on 14887  degrees of freedom
AIC: 14640

Number of Fisher Scoring iterations: 5

                                   VIF
GSEGRD                        1.413475
LowerLimitAge                 2.227446
BLS_FEDERAL_OtherSep_Rate     1.118018
BLS_FEDERAL_JobOpenings_Level 1.080933
LOSSqrt                       1.407857
SEPCount_EFDATE_OCCLog        1.090707
AGELVL_B                      1.152181
AGELVL_C                      1.653078
AGELVL_E                      1.400079
AGELVL_F                      1.275591
AGELVL_G                      1.262241
AGELVL_H                      1.355127
AGELVL_I                      1.508955
LOC_04                        1.036355
LOC_06                        1.046846
LOC_20                        1.007621
LOC_29                        1.015550
LOC_30                        1.014367
LOC_35                        1.029377
LOC_39                        1.018538
LOC_42                        1.019846
LOC_46                        1.014850
LOC_48                        1.052041
LOC_51                        1.062111
LOC_53                        1.024737
TOA_20                        1.014067
TOA_30                        1.170833
TOA_35                        1.052718
TOA_38                        1.057918
TOA_42                        1.006482
TOA_44                        1.002333
PPGROUP_11                    1.132295

Removed AFTER this step: SEPCount_EFDATE_OCCLog 


Following variables removed based on p-values (in order of removal):
 [1] "TOA_45"                   "LOC_21"                  
 [3] "LOC_09"                   "LOC_36"                  
 [5] "LOC_54"                   "LOC_16"                  
 [7] "LOC_23"                   "LOC_45"                  
 [9] "LOC_12"                   "SalaryOverUnderIndAvg"   
[11] "LOC_05"                   "LOC_10"                  
[13] "BLS_FEDERAL_Layoffs_Rate" "LOC_31"                  
[15] "LOC_33"                   "LOC_37"                  
[17] "LOC_26"                   "LOC_15"                  
[19] "LOC_41"                   "TOA_32"                  
[21] "LOC_28"                   "LOC_22"                  
[23] "TOA_40"                   "LOC_02"                  
[25] "LOC_19"                   "LOC_44"                  
[27] "LOC_50"                   "LOC_25"                  
[29] "LOC_32"                   "LOC_40"                  
[31] "LOC_08"                   "TOA_15"                  
[33] "LOC_27"                   "LOC_18"                  
[35] "LOC_47"                   "LOC_49"                  
[37] "LOC_34"                   "LOC_38"                  
[39] "LOC_13"                   "LOC_55"                  
[41] "BLS_FEDERAL_Quits_Rate"   "LOC_01"                  
[43] "LOC_24"                   "LOC_17"                  
[45] "SEPCount_EFDATE_OCCLog"  


Null Deviances (in order):
 [1] "6140.09618413702" "6139.70530359788" "6139.68430874922" "6139.65854475106"
 [5] "6139.63313569129" "6139.59532196834" "6139.53772461692" "6139.48366441678"
 [9] "6139.39728155705" "6139.29830098624" "6139.1900882605"  "6139.08688894165"
[13] "6138.98092947028" "6138.85674008882" "6138.72579420097" "6138.59742406333"
[17] "6138.47085903648" "6138.27554988296" "6138.0311195164"  "6137.73506640671"
[21] "6137.4303765745"  "6137.10477253365" "6136.75626463363" "6136.40051668325"
[25] "6136.01092268138" "6135.58921284231" "6135.16054261118" "6134.63328743168"
[29] "6134.04181569797" "6133.25422902175" "6132.31543711746" "6131.41750130868"
[33] "6130.31513080467" "6129.17306670617" "6127.97760516643" "6126.80371644396"
[37] "6125.66955201923" "6124.55307944912" "6122.89611809423" "6121.27266108516"
[41] "6119.17124816164" "6117.1107132246"  "6114.87372224233" "6112.02966681198"
[45] "6109.40921243411"

Min value at iteration =  45

Diff Degrees of Freedom (in order):
 [1] "76" "75" "74" "73" "72" "71" "70" "69" "68" "67" "66" "65" "64" "63" "62"
[16] "61" "60" "59" "58" "57" "56" "55" "54" "53" "52" "51" "50" "49" "48" "47"
[31] "46" "45" "44" "43" "42" "41" "40" "39" "38" "37" "36" "35" "34" "33" "32"

Min value at iteration =  45

Log Likelihoods (in order):
 [1] "-7271.52432906914" "-7271.71976933871" "-7271.73026676304"
 [4] "-7271.74314876212" "-7271.75585329201" "-7271.77476015348"
 [7] "-7271.80355882919" "-7271.83058892926" "-7271.87378035912"
[10] "-7271.92327064453" "-7271.9773770074"  "-7272.02897666682"
[13] "-7272.08195640251" "-7272.14405109324" "-7272.20952403716"
[16] "-7272.27370910599" "-7272.33699161941" "-7272.43464619617"
[19] "-7272.55686137945" "-7272.70488793429" "-7272.8572328504" 
[22] "-7273.02003487083" "-7273.19428882084" "-7273.37216279602"
[25] "-7273.56695979696" "-7273.7778147165"  "-7273.99214983206"
[28] "-7274.25577742181" "-7274.55151328866" "-7274.94530662677"
[31] "-7275.41470257892" "-7275.86367048331" "-7276.41485573531"
[34] "-7276.98588778457" "-7277.58361855443" "-7278.17056291567"
[37] "-7278.73764512804" "-7279.29588141309" "-7280.12436209054"
[40] "-7280.93609059507" "-7281.98679705683" "-7283.01706452535"
[43] "-7284.13556001649" "-7285.55758773166" "-7286.8678149206" 

Min value at iteration =  45

AIC values (in order):
 [1] "14697.0486581383" "14695.4395386774" "14693.4605335261" "14691.4862975242"
 [5] "14689.511706584"  "14687.549520307"  "14685.6071176584" "14683.6611778585"
 [9] "14681.7475607182" "14679.8465412891" "14677.9547540148" "14676.0579533336"
[13] "14674.163912805"  "14672.2881021865" "14670.4190480743" "14668.547418212" 
[17] "14666.6739832388" "14664.8692923923" "14663.1137227589" "14661.4097758686"
[21] "14659.7144657008" "14658.0400697417" "14656.3885776417" "14654.744325592" 
[25] "14653.1339195939" "14651.555629433"  "14649.9842996641" "14648.5115548436"
[29] "14647.1030265773" "14645.8906132535" "14644.8294051578" "14643.7273409666"
[33] "14642.8297114706" "14641.9717755691" "14641.1672371089" "14640.3411258313"
[37] "14639.4752902561" "14638.5917628262" "14638.2487241811" "14637.8721811901"
[41] "14637.9735941137" "14638.0341290507" "14638.271120033"  "14639.1151754633"
[45] "14639.7356298412"

Min value at iteration =  40

BIC values (in order):
 [1] "15283.0539144176" "15273.834337083"  "15264.2448740579" "15254.6601801823"
 [5] "15245.0751313683" "15235.5024872175" "15225.9496266952" "15216.3932290216"
 [9] "15206.8691540075" "15197.3576767046" "15187.8554315566" "15178.3481730017"
[13] "15168.8436745993" "15159.357406107"  "15149.8778941211" "15140.395806385" 
[17] "15130.911913538"  "15121.4967648178" "15112.1307373106" "15102.8163325465"
[21] "15093.510564505"  "15084.2257106721" "15074.9637606983" "15065.709050775" 
[25] "15056.4881869031" "15047.2994388684" "15038.1176512258" "15029.0344485315"
[29] "15020.0154623915" "15011.1925911939" "15002.5209252245" "14993.8084031595"
[33] "14985.3003157897" "14976.8319220145" "14968.4169256805" "14959.9803565292"
[37] "14951.5040630801" "14943.0100777765" "14935.0565812576" "14927.0695803929"
[41] "14919.5605354427" "14912.010612506"  "14904.6371456145" "14897.8707431711"
[45] "14890.8807396752"

Min value at iteration =  45

%R LRSigCols <- varsP.Repeat

array(['GSEGRD', 'LowerLimitAge', 'BLS_FEDERAL_OtherSep_Rate',
       'BLS_FEDERAL_JobOpenings_Level', 'LOSSqrt', 'AGELVL_B', 'AGELVL_C',
       'AGELVL_E', 'AGELVL_F', 'AGELVL_G', 'AGELVL_H', 'AGELVL_I',
       'LOC_04', 'LOC_06', 'LOC_20', 'LOC_29', 'LOC_30', 'LOC_35',
       'LOC_39', 'LOC_42', 'LOC_46', 'LOC_48', 'LOC_51', 'LOC_53',
       'TOA_20', 'TOA_30', 'TOA_35', 'TOA_38', 'TOA_42', 'TOA_44',
       'PPGROUP_11'], 
      dtype='<U29')

%R -o LRSigCols

#from rpy2.robjects import pandas2ri
#pandas2ri.activate()

#LRSigCols
LRSigCols = pandas2ri.ri2py(LRSigCols)
LRSigCols = LRSigCols.tolist()

%%time

def lr_explorBinary(  cost,
                      Data        = OPMAnalysisDataNoFamBinary,
                      cols        = PCList,
                      cv          = cv,
                      seed        = seed):
    
    startTime = datetime.now()
    y = Data["SEP"].values # get the labels we want    
    
    if ("SEP" in cols):    X = Data[cols].drop("SEP", axis=1).as_matrix() 
    else: X = Data[cols]
    
    lr_clf = LogisticRegression(penalty='l2', C=cost, class_weight=None, random_state=seed) # get object
    
    # setup pipeline to take PCA, then fit a clf model
    clf_pipe = Pipeline(
        [('minMaxScaler',MinMaxScaler()),
         ('CLF',lr_clf)]
    )

    accuracy = cross_val_score(clf_pipe, X, y, cv=cv.split(X, y)) # this also can help with parallelism
    MeanAccuracy =  sum(accuracy)/len(accuracy)
    accuracy = np.append(accuracy, MeanAccuracy)
    endTime = datetime.now()
    TotalTime = endTime - startTime
    accuracy = np.append(accuracy, TotalTime)
    return accuracy

CPU times: user 5 µs, sys: 0 ns, total: 5 µs
Wall time: 11.7 µs

%%time

def lr_explorBinary_w_PCA(cost,
                          PCA,
                          Data        = OPMAnalysisDataNoFamBinary,
                          cv          = cv,
                          seed        = seed):
    
    startTime = datetime.now()
    y = Data["SEP"].values # get the labels we want    
    
    X = Data.drop("SEP", axis=1).as_matrix() 
    
    lr_clf = LogisticRegression(penalty='l2', C=cost, class_weight=None, random_state=seed) # get object
    
    # setup pipeline to take PCA, then fit a clf model
    clf_pipe = Pipeline(
        [('minMaxScaler',MinMaxScaler()),
         ('PCA',PCA),
         ('CLF',lr_clf)]
    )

    accuracy = cross_val_score(clf_pipe, X, y, cv=cv.split(X, y)) # this also can help with parallelism
    MeanAccuracy =  sum(accuracy)/len(accuracy)
    accuracy = np.append(accuracy, MeanAccuracy)
    endTime = datetime.now()
    TotalTime = endTime - startTime
    accuracy = np.append(accuracy, TotalTime)
    return accuracy

CPU times: user 5 µs, sys: 0 ns, total: 5 µs
Wall time: 9.06 µs

%%time

##Full Columns
acclist = [] 

cost    = [.00000001, .0001, .001, .01, .05, 1.0, 2.0, 3.0, 4.0, 5.0]

for i in range(0,len(cost)):
    acclist.append(lr_explorBinary(cost       = cost[i],
                             cols       = fullColumns))

LRdf = pd.DataFrame(pd.concat([pd.DataFrame({
                                                "ModelVersion": "Logistic Regression: All Raw Features",
                                                "Cost": cost
                                              })[["ModelVersion", "Cost"]],
                               pd.DataFrame(acclist)], axis = 1).reindex())
LRdf.columns = ['ModelVersion','Cost', 'Iteration 0', 'Iteration 1', 'Iteration 2', 'Iteration 3', 'Iteration 4', 'MeanAccuracy', 'RunTime']
display(LRdf)
TopResultsDF = pd.concat([TopResultsDF, LRdf.sort_values(['MeanAccuracy'], ascending=False)[TopResultsDF.columns].head(1)]).sort_values(['MeanAccuracy'], ascending=False).reset_index(drop=True)
del LRdf, acclist

##Reduced Columns
acclist = [] 

cost    = [.00000001, .0001, .001, .01, .05, 1.0, 2.0, 3.0, 4.0, 5.0]

for i in range(0,len(cost)):
    acclist.append(lr_explorBinary(cost       = cost[i]))

LRdf = pd.DataFrame(pd.concat([pd.DataFrame({
                                                "ModelVersion": "Logistic Regression: Top 15 from PCA Raw Features",
                                                "Cost": cost
                                              })[["ModelVersion", "Cost"]],
                               pd.DataFrame(acclist)], axis = 1).reindex())
LRdf.columns = ['ModelVersion','Cost', 'Iteration 0', 'Iteration 1', 'Iteration 2', 'Iteration 3', 'Iteration 4', 'MeanAccuracy', 'RunTime']
display(LRdf)
TopResultsDF = pd.concat([TopResultsDF, LRdf.sort_values(['MeanAccuracy'], ascending=False)[TopResultsDF.columns].head(1)]).sort_values(['MeanAccuracy'], ascending=False).reset_index(drop=True)
del LRdf, acclist

##With PCA
acclist = [] 

cost    = [.00000001, .0001, .001, .01, .05, 1.0, 2.0, 3.0, 4.0, 5.0]

for i in range(0,len(cost)):
    acclist.append(lr_explorBinary_w_PCA(cost       = cost[i],
                                   PCA        = PCA(n_components=23, svd_solver='randomized', random_state = seed)))

LRdf = pd.DataFrame(pd.concat([pd.DataFrame({
                                                "ModelVersion": "Logistic Regression: With PCA",
                                                "Cost": cost
                                              })[["ModelVersion", "Cost"]],
                               pd.DataFrame(acclist)], axis = 1).reindex())
LRdf.columns = ['ModelVersion','Cost', 'Iteration 0', 'Iteration 1', 'Iteration 2', 'Iteration 3', 'Iteration 4', 'MeanAccuracy', 'RunTime']
display(LRdf)
TopResultsDF = pd.concat([TopResultsDF, LRdf.sort_values(['MeanAccuracy'], ascending=False)[TopResultsDF.columns].head(1)]).sort_values(['MeanAccuracy'], ascending=False).reset_index(drop=True)
del LRdf, acclist


##Significant Column List from Manual Tuning in R
acclist = [] 

cost    = [.00000001, .0001, .001, .01, .05, 1.0, 2.0, 3.0, 4.0, 5.0]

for i in range(0,len(cost)):
    acclist.append(lr_explorBinary(cost       = cost[i],
                             cols       = LRSigCols))

LRdf = pd.DataFrame(pd.concat([pd.DataFrame({
                                                "ModelVersion": "Logistic Regression: Manual Significant Features",
                                                "Cost": cost
                                              })[["ModelVersion", "Cost"]],
                               pd.DataFrame(acclist)], axis = 1).reindex())
LRdf.columns = ['ModelVersion','Cost', 'Iteration 0', 'Iteration 1', 'Iteration 2', 'Iteration 3', 'Iteration 4', 'MeanAccuracy', 'RunTime']
display(LRdf)
TopResultsDF = pd.concat([TopResultsDF, LRdf.sort_values(['MeanAccuracy'], ascending=False)[TopResultsDF.columns].head(1)]).sort_values(['MeanAccuracy'], ascending=False).reset_index(drop=True)
del LRdf, acclist

CPU times: user 2min 39s, sys: 14min 46s, total: 17min 26s
Wall time: 28.7 s

display(TopResultsDF)

plot = TopResultsDF[["Iteration 0","Iteration 1","Iteration 2","Iteration 3","Iteration 4"]].transpose().plot.line(title = "Top Results Among Varying Model Feature Inputs",rot=45)
plot.set_xlabel("Iterations")
plot.set_ylabel("Accuracies")
plot.legend(loc='center left', bbox_to_anchor=(1.01, .5))

FinalResultsDF = pd.concat([FinalResultsDF, TopResultsDF.sort_values(['MeanAccuracy'], ascending=False)[TopResultsDF.columns].head(1)]).sort_values(['MeanAccuracy'], ascending=False).reset_index(drop=True)
TopResultsDF = pd.DataFrame(columns= ['ModelVersion', 'Iteration 0', 'Iteration 1', 'Iteration 2', 'Iteration 3', 'Iteration 4', 'MeanAccuracy'])

/usr/local/es7/lib/python3.5/site-packages/matplotlib/font_manager.py:1297: UserWarning: findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans
  (prop.get_family(), self.defaultFamily[fontext]))

lr_clf = LogisticRegression(penalty='l2', C=1, class_weight=None, random_state=seed) # get object

lr_clf, lr_acc = compute_kfold_scores_ClassificationBinary(clf         = lr_clf,
                                                           cols        = LRSigCols)

/usr/local/es7/lib/python3.5/site-packages/matplotlib/font_manager.py:1297: UserWarning: findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans
  (prop.get_family(), self.defaultFamily[fontext]))

confusion matrix
Predicted    NS    SC   All
True                       
NS          986   514  1500
SC          435  1050  1485
All        1421  1564  2985

Normalized confusion matrix
[[ 0.65733333  0.34266667]
 [ 0.29292929  0.70707071]]

/usr/local/es7/lib/python3.5/site-packages/matplotlib/font_manager.py:1297: UserWarning: findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans
  (prop.get_family(), self.defaultFamily[fontext]))

confusion matrix
Predicted    NS    SC   All
True                       
NS         1206   294  1500
SC          297  1188  1485
All        1503  1482  2985

Normalized confusion matrix
[[ 0.804  0.196]
 [ 0.2    0.8  ]]

/usr/local/es7/lib/python3.5/site-packages/matplotlib/font_manager.py:1297: UserWarning: findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans
  (prop.get_family(), self.defaultFamily[fontext]))

confusion matrix
Predicted   NS    SC   All
True                      
NS         815   684  1499
SC         115  1370  1485
All        930  2054  2984

Normalized confusion matrix
[[ 0.5436958   0.4563042 ]
 [ 0.07744108  0.92255892]]

/usr/local/es7/lib/python3.5/site-packages/matplotlib/font_manager.py:1297: UserWarning: findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans
  (prop.get_family(), self.defaultFamily[fontext]))

confusion matrix
Predicted    NS    SC   All
True                       
NS         1046   453  1499
SC          270  1214  1484
All        1316  1667  2983

Normalized confusion matrix
[[ 0.69779853  0.30220147]
 [ 0.1819407   0.8180593 ]]

/usr/local/es7/lib/python3.5/site-packages/matplotlib/font_manager.py:1297: UserWarning: findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans
  (prop.get_family(), self.defaultFamily[fontext]))

confusion matrix
Predicted    NS    SC   All
True                       
NS         1052   447  1499
SC          305  1179  1484
All        1357  1626  2983

Normalized confusion matrix
[[ 0.7018012   0.2981988 ]
 [ 0.20552561  0.79447439]]

Retest Random Forest and KNN with Logistic Regression Significant Columns¶

#### Random Forest Sign. Cols

acclist = []

n_estimators       =  [10    , 10     , 10    , 10    , 10    , 10    , 10    , 10    , 10    , 10   , 10   , 10    , 10    , 10    , 10    , 10    , 10    , 10    , 10    , 10    , 10    , 10    , 10    , 10    , 10    , 10    , 10    , 15    , 20    , 30    , 50    ]  
max_features       =  ['auto', 'auto' , 'auto', 'auto', 'auto', 'auto', 'auto', 5     , 10    , 15   , 20   , None  , 5     , 5     , 5     , 5     , 5     , 5     , 5     , 5     , 5     , 5     , 5     , 5     , 5     , 5     , 5     , 5     , 5     , 5     , 5     ] 
max_depth          =  [None  , None   , None  , None  , None  , None  , None  , None  , None  , None , None , None  , 10    , 15    , 20    , 25    , 30    , 3     , 4     , 5     , 6     , 7     , 8     , 9     , 11    , 12    , 13    , 4     , 4     , 4     , 4     ] 
min_samples_split  =  [2     , 8      , 12    , 18    , 20    , 24    , 36    , 18    , 18    , 18   , 18   , 18    , 18    , 18    , 18    , 18    , 18    , 18    , 18    , 18    , 18    , 18    , 18    , 18    , 18    , 18    , 18    , 18    , 18    , 18    , 18    ] 
min_samples_leaf   =  [1     , 4      , 6     , 9     , 10    , 12    , 18    , 9     , 9     , 9    , 9    , 9     , 9     , 9     , 9     , 9     , 9     , 9     , 9     , 9     , 9     , 9     , 9     , 9     , 9     , 9     , 9     , 9     , 9     , 9     , 9     ]


## Model with only top 15 raw Scaled Principal Features 
for i in range(0,len(n_estimators)):
    acclist.append(rfc_explorBinary(n_estimators      = n_estimators[i],
                              max_features      = max_features[i],
                              max_depth         = max_depth[i],
                              min_samples_split = min_samples_split[i],
                              min_samples_leaf  = min_samples_leaf[i],
                              cols = LRSigCols
                             )
                  )

rfcdf = pd.DataFrame(pd.concat([pd.DataFrame({  "ModelVersion": "Random Forest: With LR Sig Cols",
                                                "n_estimators": n_estimators,          
                                                "max_features": max_features,         
                                                "max_depth": max_depth,        
                                                "min_samples_split": min_samples_split,
                                                "min_samples_leaf": min_samples_leaf   
                                              }),
                               pd.DataFrame(acclist)], axis = 1).reindex())
rfcdf.columns = ['ModelVersion', 'max_depth', 'max_features', 'min_samples_leaf','min_samples_split', 'n_estimators', 'Iteration 0', 'Iteration 1', 'Iteration 2', 'Iteration 3', 'Iteration 4', 'MeanAccuracy', 'RunTime']
display(rfcdf)
TopResultsDF = pd.concat([TopResultsDF, rfcdf.sort_values(['MeanAccuracy'], ascending=False)[TopResultsDF.columns].head(1)]).sort_values(['MeanAccuracy'], ascending=False).reset_index(drop=True)
del rfcdf, acclist


#### KNN Sign. Cols

acclist = [] 

n_neighbors =  [5          , 10         , 15         , 20         , 30         , 40         , 50         , 100        , 150        , 200        , 250        , 200        , 200        , 200        , 200        , 200        , 200        , 200     , 200]
algorithm   =  'ball_tree'
leaf_size   =  [30         , 30         , 30         , 30         , 30         , 30         , 30         , 30         , 30         , 30         , 30         , 2          , 3          , 4          , 5         , 20        , 50          , 100      , 150]



for i in range(0,len(n_neighbors)):
    acclist.append(knn_explorBinary(n_neighbors = n_neighbors[i],
                              algorithm   = algorithm,
                              leaf_size   = leaf_size[i],
                              cols = LRSigCols
                             )
                  )

rfcdf = pd.DataFrame(pd.concat([pd.DataFrame({
                                                "ModelVersion": "KNN: " + algorithm + ", With LR Sig Cols",
                                                "n_neighbors": n_neighbors,          
                                                "algorithm": algorithm,         
                                                "leaf_size": leaf_size  
                                              }),
                               pd.DataFrame(acclist)], axis = 1).reindex())
rfcdf.columns = ['ModelVersion','algorithm', 'leaf_size','n_neighbors', 'Iteration 0', 'Iteration 1', 'Iteration 2', 'Iteration 3', 'Iteration 4', 'MeanAccuracy', 'RunTime']
display(rfcdf)
TopResultsDF = pd.concat([TopResultsDF, rfcdf.sort_values(['MeanAccuracy'], ascending=False)[TopResultsDF.columns].head(1)]).sort_values(['MeanAccuracy'], ascending=False).reset_index(drop=True)
del rfcdf, acclist


acclist = [] 

n_neighbors =  [5          , 10         , 15         , 20         , 30         , 40         , 50         , 100        , 150        , 200        , 250        , 200        , 200        , 200        , 200        , 200        , 200        , 200     , 200]
algorithm   =  'kd_tree'
leaf_size   =  [30         , 30         , 30         , 30         , 30         , 30         , 30         , 30         , 30         , 30         , 30         , 2          , 3          , 4          , 5         , 20        , 50          , 100      , 150]



for i in range(0,len(n_neighbors)):
    acclist.append(knn_explorBinary(n_neighbors = n_neighbors[i],
                              algorithm   = algorithm,
                              leaf_size   = leaf_size[i],
                              cols = LRSigCols
                             )
                  )

rfcdf = pd.DataFrame(pd.concat([pd.DataFrame({
                                                "ModelVersion": "KNN: " + algorithm + ", With LR Sig Cols",
                                                "n_neighbors": n_neighbors,          
                                                "algorithm": algorithm,         
                                                "leaf_size": leaf_size  
                                              }),
                               pd.DataFrame(acclist)], axis = 1).reindex())
rfcdf.columns = ['ModelVersion','algorithm', 'leaf_size','n_neighbors', 'Iteration 0', 'Iteration 1', 'Iteration 2', 'Iteration 3', 'Iteration 4', 'MeanAccuracy', 'RunTime']
display(rfcdf)
TopResultsDF = pd.concat([TopResultsDF, rfcdf.sort_values(['MeanAccuracy'], ascending=False)[TopResultsDF.columns].head(1)]).sort_values(['MeanAccuracy'], ascending=False).reset_index(drop=True)
del rfcdf, acclist

display(TopResultsDF)

plot = TopResultsDF[["Iteration 0","Iteration 1","Iteration 2","Iteration 3","Iteration 4"]].transpose().plot.line(title = "Top Results Among Varying Model Feature Inputs",rot=45)
plot.set_xlabel("Iterations")
plot.set_ylabel("Accuracies")
plot.legend(loc='center left', bbox_to_anchor=(1.01, .5))

FinalResultsDF = pd.concat([FinalResultsDF, TopResultsDF.sort_values(['MeanAccuracy'], ascending=False)[TopResultsDF.columns].head(1)]).sort_values(['MeanAccuracy'], ascending=False).reset_index(drop=True)
TopResultsDF = pd.DataFrame(columns= ['ModelVersion', 'Iteration 0', 'Iteration 1', 'Iteration 2', 'Iteration 3', 'Iteration 4', 'MeanAccuracy'])

/usr/local/es7/lib/python3.5/site-packages/matplotlib/font_manager.py:1297: UserWarning: findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans
  (prop.get_family(), self.defaultFamily[fontext]))

%%time

rfc_clf = RandomForestClassifier(n_estimators       =50, 
                                 max_features       = 5, 
                                 max_depth          = 4.0, 
                                 min_samples_split  = 18, 
                                 min_samples_leaf   = 9,
                                 n_jobs             = -1, 
                                 random_state       = seed) # get object
    
rfc_clf, rfc_acc = compute_kfold_scores_ClassificationBinary(rfc_clf, 
                                                             ##PCA = PCA(n_components=22, svd_solver='randomized', random_state = seed),
                                                             cols = LRSigCols)

/usr/local/es7/lib/python3.5/site-packages/matplotlib/font_manager.py:1297: UserWarning: findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans
  (prop.get_family(), self.defaultFamily[fontext]))

confusion matrix
Predicted    NS    SC   All
True                       
NS          871   629  1500
SC          215  1270  1485
All        1086  1899  2985

Normalized confusion matrix
[[ 0.58066667  0.41933333]
 [ 0.14478114  0.85521886]]

/usr/local/es7/lib/python3.5/site-packages/matplotlib/font_manager.py:1297: UserWarning: findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans
  (prop.get_family(), self.defaultFamily[fontext]))

confusion matrix
Predicted    NS    SC   All
True                       
NS         1296   204  1500
SC          536   949  1485
All        1832  1153  2985

Normalized confusion matrix
[[ 0.864       0.136     ]
 [ 0.36094276  0.63905724]]

/usr/local/es7/lib/python3.5/site-packages/matplotlib/font_manager.py:1297: UserWarning: findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans
  (prop.get_family(), self.defaultFamily[fontext]))

confusion matrix
Predicted   NS    SC   All
True                      
NS         838   661  1499
SC         147  1338  1485
All        985  1999  2984

Normalized confusion matrix
[[ 0.55903936  0.44096064]
 [ 0.0989899   0.9010101 ]]

/usr/local/es7/lib/python3.5/site-packages/matplotlib/font_manager.py:1297: UserWarning: findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans
  (prop.get_family(), self.defaultFamily[fontext]))

confusion matrix
Predicted    NS    SC   All
True                       
NS          951   548  1499
SC          191  1293  1484
All        1142  1841  2983

Normalized confusion matrix
[[ 0.63442295  0.36557705]
 [ 0.1287062   0.8712938 ]]

/usr/local/es7/lib/python3.5/site-packages/matplotlib/font_manager.py:1297: UserWarning: findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans
  (prop.get_family(), self.defaultFamily[fontext]))

confusion matrix
Predicted    NS    SC   All
True                       
NS          901   598  1499
SC          228  1256  1484
All        1129  1854  2983

Normalized confusion matrix
[[ 0.60106738  0.39893262]
 [ 0.15363881  0.84636119]]

%%time
import os
from sklearn import tree
import pydotplus
import six
from sklearn.tree import export_graphviz
from IPython.display import SVG 



i_tree = 0
for tree_in_forest in rfc_clf.estimators_:       
    svgData = tree.export_graphviz(tree_in_forest, 
                         feature_names=fullColumns,
                         class_names=["NS", "SD"],
                         filled=True,
                         #rounded=True,
                         rotate = True,
                         label = 'All',
                         out_file=None)


    graph=pydotplus.graph_from_dot_data(svgData)
    
    if not os.path.exists('images'):
        os.makedirs('images')
    
    graph.write_svg('images/tree'+ str(i_tree) +'.svg')
        
    i_tree = i_tree + 1

CPU times: user 11.7 s, sys: 1.69 s, total: 13.3 s
Wall time: 16.6 s

SVG(filename='images/tree0.svg')

display(FinalResultsDF)
#plot top results across all model tests 
plot = FinalResultsDF[["Iteration 0","Iteration 1","Iteration 2","Iteration 3","Iteration 4"]].transpose().plot.line(title = "Final Results across Model Types",rot=45)
plot.set_xlabel("Iterations")
plot.set_ylabel("Accuracies")
plot.legend(loc='center left', bbox_to_anchor=(1.01, .5))
del FinalResultsDF

/usr/local/es7/lib/python3.5/site-packages/matplotlib/font_manager.py:1297: UserWarning: findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans
  (prop.get_family(), self.defaultFamily[fontext]))

Feature Importance of fit models¶

Binary Classifier Logistic Regression¶

print(lr_clf.coef_[0])
coef = pd.Series(lr_clf.coef_[0], index=LRSigCols)

maxcoef = pd.Series(pd.DataFrame(abs(coef).sort_values(ascending=False).head(20)).index)
       
weightsplot = pd.Series(coef, index=maxcoef)
weightsplot.plot(title = "Logistic Regression Coefficients", kind='bar', color = 'Tomato')

[-0.53761554 -2.0345517   0.76432449  0.39580152 -6.3888843  -0.48416436
 -0.12829652  0.31087776  0.51766391  0.74976631  0.86371825  0.67961665
  0.55969504  0.4690265   0.44315626 -0.39532162  0.69560419  0.28090379
 -0.20428945 -0.26813639  0.80055387  0.29524038 -0.07782224  0.37876725
  0.84198517  0.2849319  -0.60110599  0.35630518  1.09238682  1.08448602
 -0.46266585]

<matplotlib.axes._subplots.AxesSubplot at 0x7f7227bebb00>

/usr/local/es7/lib/python3.5/site-packages/matplotlib/font_manager.py:1297: UserWarning: findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans
  (prop.get_family(), self.defaultFamily[fontext]))

Define Model Fit on full train Professional Data¶

Using the full datset, to create our model fit allows us to fully utilize our dataset instead of simply utilizing the last 80% training fold fit on external data.

y = OPMAnalysisDataNoFamBinary["SEP"].values # get the labels we want    
y = np.where(y=="NS",0,1) # turn into numeric binary
X = OPMAnalysisDataNoFamBinary.drop("SEP", axis=1)

XFC = pd.DataFrame(OPMAnalysisScalerFit.transform(X),columns=X.columns)[fullColumns].as_matrix() 
XPCC = pd.DataFrame(OPMAnalysisScalerFit.transform(X),columns=X.columns)[PCList]#.as_matrix() 
XSigC = pd.DataFrame(OPMAnalysisScalerFit.transform(X),columns=X.columns)[LRSigCols].as_matrix()

rfc_clf = RandomForestClassifier(n_estimators       =50, 
                                 max_features       = 5, 
                                 max_depth          = 4.0, 
                                 min_samples_split  = 18, 
                                 min_samples_leaf   = 9,
                                 n_jobs             = -1, 
                                 random_state       = seed) # get object

rfc_clf.fit(XSigC,y)

RandomForestClassifier(bootstrap=True, class_weight=None, criterion='gini',
            max_depth=4.0, max_features=5, max_leaf_nodes=None,
            min_impurity_split=1e-07, min_samples_leaf=9,
            min_samples_split=18, min_weight_fraction_leaf=0.0,
            n_estimators=50, n_jobs=-1, oob_score=False,
            random_state=14920, verbose=0, warm_start=False)

knn_clf = KNeighborsClassifier(n_neighbors = 250, algorithm = 'kd_tree',leaf_size = 30, n_jobs=-1) # get object

knn_clf.fit(XPCC,y)

KNeighborsClassifier(algorithm='kd_tree', leaf_size=30, metric='minkowski',
           metric_params=None, n_jobs=-1, n_neighbors=250, p=2,
           weights='uniform')

lr_clf = LogisticRegression(penalty='l2', C=1, class_weight=None, random_state=seed) # get object

lr_clf.fit(XSigC,y)

LogisticRegression(C=1, class_weight=None, dual=False, fit_intercept=True,
          intercept_scaling=1, max_iter=100, multi_class='ovr', n_jobs=1,
          penalty='l2', random_state=14920, solver='liblinear', tol=0.0001,
          verbose=0, warm_start=False)

Predicting Admin from the Professional Model¶

%%time

if os.path.isfile(PickleJarPath+"/OPMAnalysisDataNoFamAdminBinary.pkl"):
    print("Found the File! Loading Pickle Now!")
    OPMAnalysisDataNoFamAdminBinary = unpickleObject("OPMAnalysisDataNoFamAdminBinary")
else:

    OPMAnalysisDataNoFamAdminBinary = SampledOPMDataAdmin.copy()

    cols = ["GENDER",
            "DATECODE",
            "QTR",
            "COUNT",
            "AGYTYPT",
            "AGYT",
            "AGYSUB",
            "AGYSUBT",
            "QTR",
            "AGELVLT",
            "LOSLVL",
            "LOSLVLT",
            "LOCTYPT",
            "LOCT",
            "OCCTYP",
            "OCCTYPT",
            "OCCFAM",
            "OCCFAMT",
            "OCC",
            "OCCT",
            "PATCO",
            "PPGRD",
            "PATCOT",
            "PPTYPT",
            "PPGROUPT",
            "PAYPLAN",
            "PAYPLANT",
            "SALLVLT",
            "TOATYPT",
            "TOAT",
            "WSTYP",
            "WSTYPT",
            "WORKSCH",
            "WORKSCHT",
            "SALARY",
            "LOS",
            "SEPCount_EFDATE_OCC",
            "SEPCount_EFDATE_LOC"
           ]



    #delete cols from analysis data
    for col in cols:
        if col in list(OPMAnalysisDataNoFamAdminBinary.columns):
            del OPMAnalysisDataNoFamAdminBinary[col]

    OPMAnalysisDataNoFamAdminBinary.info()

    cols = ["AGELVL",
            "LOC",
            "SALLVL",
            "TOA",
            "AGYTYP",
            "AGY",
            "LOCTYP",
            "PPTYP",
            "PPGROUP",
            "TOATYP"
           ]

    #Split Values for cols 
    for col in cols:
        if col in list(OPMAnalysisDataNoFamAdminBinary.columns):
            AttSplit = pd.get_dummies(OPMAnalysisDataNoFamAdminBinary[col],prefix=col)
            display(AttSplit.head())
            OPMAnalysisDataNoFamAdminBinary = pd.concat((OPMAnalysisDataNoFamAdminBinary,AttSplit),axis=1) # add back into the dataframe
            del OPMAnalysisDataNoFamAdminBinary[col]

    pickleObject(OPMAnalysisDataNoFamAdminBinary, "OPMAnalysisDataNoFamAdminBinary")
        
display(OPMAnalysisDataNoFamAdminBinary.head())
print("Number of Columns: ",len(OPMAnalysisDataNoFamAdminBinary.columns))
OPMAnalysisDataNoFamAdminBinary.info()

Found the File! Loading Pickle Now!

Number of Columns:  100
<class 'pandas.core.frame.DataFrame'>
RangeIndex: 14918 entries, 0 to 14917
Data columns (total 100 columns):
SEP                              14918 non-null object
GSEGRD                           14918 non-null float64
IndAvgSalary                     14918 non-null float64
SalaryOverUnderIndAvg            14918 non-null float64
LowerLimitAge                    14918 non-null float64
YearsToRetirement                14918 non-null float64
BLS_FEDERAL_OtherSep_Rate        14918 non-null float64
BLS_FEDERAL_Quits_Rate           14918 non-null float64
BLS_FEDERAL_TotalSep_Level       14918 non-null int64
BLS_FEDERAL_JobOpenings_Rate     14918 non-null float64
BLS_FEDERAL_OtherSep_Level       14918 non-null int64
BLS_FEDERAL_Quits_Level          14918 non-null int64
BLS_FEDERAL_JobOpenings_Level    14918 non-null int64
BLS_FEDERAL_Layoffs_Rate         14918 non-null float64
BLS_FEDERAL_Layoffs_Level        14918 non-null int64
BLS_FEDERAL_TotalSep_Rate        14918 non-null float64
SALARYLog                        14918 non-null float64
LOSSqrt                          14918 non-null float64
SEPCount_EFDATE_OCCLog           14918 non-null float64
SEPCount_EFDATE_LOCLog           14918 non-null float64
IndAvgSalaryLog                  14918 non-null float64
AGELVL_B                         14918 non-null uint8
AGELVL_C                         14918 non-null uint8
AGELVL_D                         14918 non-null uint8
AGELVL_E                         14918 non-null uint8
AGELVL_F                         14918 non-null uint8
AGELVL_G                         14918 non-null uint8
AGELVL_H                         14918 non-null uint8
AGELVL_I                         14918 non-null uint8
AGELVL_J                         14918 non-null uint8
AGELVL_K                         14918 non-null uint8
LOC_01                           14918 non-null uint8
LOC_02                           14918 non-null uint8
LOC_04                           14918 non-null uint8
LOC_05                           14918 non-null uint8
LOC_06                           14918 non-null uint8
LOC_08                           14918 non-null uint8
LOC_09                           14918 non-null uint8
LOC_10                           14918 non-null uint8
LOC_11                           14918 non-null uint8
LOC_12                           14918 non-null uint8
LOC_13                           14918 non-null uint8
LOC_15                           14918 non-null uint8
LOC_16                           14918 non-null uint8
LOC_17                           14918 non-null uint8
LOC_18                           14918 non-null uint8
LOC_19                           14918 non-null uint8
LOC_20                           14918 non-null uint8
LOC_21                           14918 non-null uint8
LOC_22                           14918 non-null uint8
LOC_23                           14918 non-null uint8
LOC_24                           14918 non-null uint8
LOC_25                           14918 non-null uint8
LOC_26                           14918 non-null uint8
LOC_27                           14918 non-null uint8
LOC_28                           14918 non-null uint8
LOC_29                           14918 non-null uint8
LOC_30                           14918 non-null uint8
LOC_31                           14918 non-null uint8
LOC_32                           14918 non-null uint8
LOC_33                           14918 non-null uint8
LOC_34                           14918 non-null uint8
LOC_35                           14918 non-null uint8
LOC_36                           14918 non-null uint8
LOC_37                           14918 non-null uint8
LOC_38                           14918 non-null uint8
LOC_39                           14918 non-null uint8
LOC_40                           14918 non-null uint8
LOC_41                           14918 non-null uint8
LOC_42                           14918 non-null uint8
LOC_44                           14918 non-null uint8
LOC_45                           14918 non-null uint8
LOC_46                           14918 non-null uint8
LOC_47                           14918 non-null uint8
LOC_48                           14918 non-null uint8
LOC_49                           14918 non-null uint8
LOC_50                           14918 non-null uint8
LOC_51                           14918 non-null uint8
LOC_53                           14918 non-null uint8
LOC_54                           14918 non-null uint8
LOC_55                           14918 non-null uint8
LOC_56                           14918 non-null uint8
TOA_10                           14918 non-null uint8
TOA_15                           14918 non-null uint8
TOA_20                           14918 non-null uint8
TOA_30                           14918 non-null uint8
TOA_32                           14918 non-null uint8
TOA_35                           14918 non-null uint8
TOA_38                           14918 non-null uint8
TOA_40                           14918 non-null uint8
TOA_42                           14918 non-null uint8
TOA_44                           14918 non-null uint8
TOA_45                           14918 non-null uint8
TOA_48                           14918 non-null uint8
LOCTYP_1                         14918 non-null uint8
PPTYP_1                          14918 non-null uint8
PPGROUP_11                       14918 non-null uint8
PPGROUP_12                       14918 non-null uint8
TOATYP_1                         14918 non-null uint8
TOATYP_2                         14918 non-null uint8
dtypes: float64(15), int64(5), object(1), uint8(79)
memory usage: 3.5+ MB
CPU times: user 94.2 ms, sys: 763 µs, total: 95 ms
Wall time: 93.4 ms

if os.path.isfile(PickleJarPath+"/OPMAnalysisScalerFit.pkl"):
    print("Found the File! Loading Pickle Now!")
    OPMAnalysisScalerFit = unpickleObject("OPMAnalysisScalerFit")

Found the File! Loading Pickle Now!

#OPMAnalysisDataNoFamAdminBinary = OPMAnalysisDataNoFamAdmin[(OPMAnalysisDataNoFamAdmin["SEP"] == 'NS') | (OPMAnalysisDataNoFamAdmin["SEP"] == 'SC')].reset_index()
OPMAnalysisDataNoFamAdminBinaryScaled = OPMAnalysisDataNoFamAdminBinary[OPMScaledAnalysisData.columns]

print(OPMAnalysisDataNoFamAdminBinaryScaled.info())

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 14918 entries, 0 to 14917
Data columns (total 99 columns):
GSEGRD                           14918 non-null float64
IndAvgSalary                     14918 non-null float64
SalaryOverUnderIndAvg            14918 non-null float64
LowerLimitAge                    14918 non-null float64
YearsToRetirement                14918 non-null float64
BLS_FEDERAL_OtherSep_Rate        14918 non-null float64
BLS_FEDERAL_Quits_Rate           14918 non-null float64
BLS_FEDERAL_TotalSep_Level       14918 non-null int64
BLS_FEDERAL_JobOpenings_Rate     14918 non-null float64
BLS_FEDERAL_OtherSep_Level       14918 non-null int64
BLS_FEDERAL_Quits_Level          14918 non-null int64
BLS_FEDERAL_JobOpenings_Level    14918 non-null int64
BLS_FEDERAL_Layoffs_Rate         14918 non-null float64
BLS_FEDERAL_Layoffs_Level        14918 non-null int64
BLS_FEDERAL_TotalSep_Rate        14918 non-null float64
SALARYLog                        14918 non-null float64
LOSSqrt                          14918 non-null float64
SEPCount_EFDATE_OCCLog           14918 non-null float64
SEPCount_EFDATE_LOCLog           14918 non-null float64
IndAvgSalaryLog                  14918 non-null float64
AGELVL_B                         14918 non-null uint8
AGELVL_C                         14918 non-null uint8
AGELVL_D                         14918 non-null uint8
AGELVL_E                         14918 non-null uint8
AGELVL_F                         14918 non-null uint8
AGELVL_G                         14918 non-null uint8
AGELVL_H                         14918 non-null uint8
AGELVL_I                         14918 non-null uint8
AGELVL_J                         14918 non-null uint8
AGELVL_K                         14918 non-null uint8
LOC_01                           14918 non-null uint8
LOC_02                           14918 non-null uint8
LOC_04                           14918 non-null uint8
LOC_05                           14918 non-null uint8
LOC_06                           14918 non-null uint8
LOC_08                           14918 non-null uint8
LOC_09                           14918 non-null uint8
LOC_10                           14918 non-null uint8
LOC_11                           14918 non-null uint8
LOC_12                           14918 non-null uint8
LOC_13                           14918 non-null uint8
LOC_15                           14918 non-null uint8
LOC_16                           14918 non-null uint8
LOC_17                           14918 non-null uint8
LOC_18                           14918 non-null uint8
LOC_19                           14918 non-null uint8
LOC_20                           14918 non-null uint8
LOC_21                           14918 non-null uint8
LOC_22                           14918 non-null uint8
LOC_23                           14918 non-null uint8
LOC_24                           14918 non-null uint8
LOC_25                           14918 non-null uint8
LOC_26                           14918 non-null uint8
LOC_27                           14918 non-null uint8
LOC_28                           14918 non-null uint8
LOC_29                           14918 non-null uint8
LOC_30                           14918 non-null uint8
LOC_31                           14918 non-null uint8
LOC_32                           14918 non-null uint8
LOC_33                           14918 non-null uint8
LOC_34                           14918 non-null uint8
LOC_35                           14918 non-null uint8
LOC_36                           14918 non-null uint8
LOC_37                           14918 non-null uint8
LOC_38                           14918 non-null uint8
LOC_39                           14918 non-null uint8
LOC_40                           14918 non-null uint8
LOC_41                           14918 non-null uint8
LOC_42                           14918 non-null uint8
LOC_44                           14918 non-null uint8
LOC_45                           14918 non-null uint8
LOC_46                           14918 non-null uint8
LOC_47                           14918 non-null uint8
LOC_48                           14918 non-null uint8
LOC_49                           14918 non-null uint8
LOC_50                           14918 non-null uint8
LOC_51                           14918 non-null uint8
LOC_53                           14918 non-null uint8
LOC_54                           14918 non-null uint8
LOC_55                           14918 non-null uint8
LOC_56                           14918 non-null uint8
TOA_10                           14918 non-null uint8
TOA_15                           14918 non-null uint8
TOA_20                           14918 non-null uint8
TOA_30                           14918 non-null uint8
TOA_32                           14918 non-null uint8
TOA_35                           14918 non-null uint8
TOA_38                           14918 non-null uint8
TOA_40                           14918 non-null uint8
TOA_42                           14918 non-null uint8
TOA_44                           14918 non-null uint8
TOA_45                           14918 non-null uint8
TOA_48                           14918 non-null uint8
LOCTYP_1                         14918 non-null uint8
PPTYP_1                          14918 non-null uint8
PPGROUP_11                       14918 non-null uint8
PPGROUP_12                       14918 non-null uint8
TOATYP_1                         14918 non-null uint8
TOATYP_2                         14918 non-null uint8
dtypes: float64(15), int64(5), uint8(79)
memory usage: 3.4 MB
None

Predicting Admin with Random Forest¶

%%time
OPMAnalysisDataNoFamAdminBinaryScaled = pd.DataFrame(OPMAnalysisScalerFit.transform(OPMAnalysisDataNoFamAdminBinaryScaled), columns = OPMAnalysisDataNoFamAdminBinaryScaled.columns)

print("Overall Accuracy, predicting Admin Binary Separation from Professional Model: ", rfc_clf.score(OPMAnalysisDataNoFamAdminBinaryScaled[LRSigCols], np.where(OPMAnalysisDataNoFamAdminBinary["SEP"]=="NS",0,1)))

results = pd.concat([OPMAnalysisDataNoFamAdminBinary, pd.DataFrame({"Prediction": rfc_clf.predict(OPMAnalysisDataNoFamAdminBinaryScaled[LRSigCols])})], axis = 1)
results["SEPNum"] = np.where(results["SEP"]=="NS",0,1)
results["PredictTxt"] = np.where(results["Prediction"]==0,"NS","SC")

display(pd.DataFrame({'Cnt' : results.groupby(["SEP"]).size()}).reset_index())
display(pd.DataFrame({'Cnt' : results.groupby(["SEP", "PredictTxt"]).size()}).reset_index())

print("confusion matrix\n{0}\n".format(pd.crosstab(results.PredictTxt, results.SEP, rownames = ['True'], colnames = ['Predicted'], margins = True)))

    # Plot non-normalized confusion matrix
plt.figure()
plot_confusion_matrixBinary(confusion_matrix(results.Prediction, results.SEPNum), 
                      classes   =["NS",  "SC"], 
                      normalize =True,
                      title     ='Confusion matrix, with normalization')

Overall Accuracy, predicting Admin Binary Separation from Professional Model:  0.726571926532

confusion matrix
Predicted    NS    SC    All
True                        
NS         4858  1442   6300
SC         2637  5981   8618
All        7495  7423  14918

Normalized confusion matrix
[[ 0.77111111  0.22888889]
 [ 0.30598747  0.69401253]]

/usr/local/es7/lib/python3.5/site-packages/matplotlib/font_manager.py:1297: UserWarning: findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans
  (prop.get_family(), self.defaultFamily[fontext]))

CPU times: user 640 ms, sys: 289 ms, total: 929 ms
Wall time: 607 ms

Predicting Admin with KNN¶

%%time

print("Overall Accuracy, predicting Admin Binary Separation from Professional Model: ", knn_clf.score(OPMAnalysisDataNoFamAdminBinaryScaled[PCList], np.where(OPMAnalysisDataNoFamAdminBinary["SEP"]=="NS",0,1)))

results = pd.concat([OPMAnalysisDataNoFamAdminBinary, pd.DataFrame({"Prediction": knn_clf.predict(OPMAnalysisDataNoFamAdminBinaryScaled[PCList])})], axis = 1)
results["SEPNum"] = np.where(results["SEP"]=="NS",0,1)
results["PredictTxt"] = np.where(results["Prediction"]==0,"NS","SC")

display(pd.DataFrame({'Cnt' : results.groupby(["SEP"]).size()}).reset_index())
display(pd.DataFrame({'Cnt' : results.groupby(["SEP", "PredictTxt"]).size()}).reset_index())

print("confusion matrix\n{0}\n".format(pd.crosstab(results.PredictTxt, results.SEP, rownames = ['True'], colnames = ['Predicted'], margins = True)))

    # Plot non-normalized confusion matrix
plt.figure()
plot_confusion_matrixBinary(confusion_matrix(results.Prediction, results.SEPNum), 
                      classes   =["NS",  "SC"], 
                      normalize =True,
                      title     ='Confusion matrix, with normalization')

Overall Accuracy, predicting Admin Binary Separation from Professional Model:  0.709880681056

confusion matrix
Predicted    NS    SC    All
True                        
NS         5554  2387   7941
SC         1941  5036   6977
All        7495  7423  14918

Normalized confusion matrix
[[ 0.69940813  0.30059187]
 [ 0.2781998   0.7218002 ]]

/usr/local/es7/lib/python3.5/site-packages/matplotlib/font_manager.py:1297: UserWarning: findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans
  (prop.get_family(), self.defaultFamily[fontext]))

CPU times: user 21min 8s, sys: 1.25 s, total: 21min 9s
Wall time: 31.3 s

Predicting Admin with Logistic Regression¶

%%time

print("Overall Accuracy, predicting Admin Binary Separation from Professional Model: ", lr_clf.score(OPMAnalysisDataNoFamAdminBinaryScaled[LRSigCols], np.where(OPMAnalysisDataNoFamAdminBinary["SEP"]=="NS",0,1)))

results = pd.concat([OPMAnalysisDataNoFamAdminBinary, pd.DataFrame({"Prediction": lr_clf.predict(OPMAnalysisDataNoFamAdminBinaryScaled[LRSigCols])})], axis = 1)
results["SEPNum"] = np.where(results["SEP"]=="NS",0,1)
results["PredictTxt"] = np.where(results["Prediction"]==0,"NS","SC")

display(pd.DataFrame({'Cnt' : results.groupby(["SEP"]).size()}).reset_index())
display(pd.DataFrame({'Cnt' : results.groupby(["SEP", "PredictTxt"]).size()}).reset_index())

print("confusion matrix\n{0}\n".format(pd.crosstab(results.PredictTxt, results.SEP, rownames = ['True'], colnames = ['Predicted'], margins = True)))

    # Plot non-normalized confusion matrix
plt.figure()
plot_confusion_matrixBinary(confusion_matrix(results.Prediction, results.SEPNum), 
                      classes   =["NS",  "SC"], 
                      normalize =True,
                      title     ='Confusion matrix, with normalization')

Overall Accuracy, predicting Admin Binary Separation from Professional Model:  0.726303794074

confusion matrix
Predicted    NS    SC    All
True                        
NS         5269  1857   7126
SC         2226  5566   7792
All        7495  7423  14918

Normalized confusion matrix
[[ 0.739405    0.260595  ]
 [ 0.28567762  0.71432238]]

/usr/local/es7/lib/python3.5/site-packages/matplotlib/font_manager.py:1297: UserWarning: findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans
  (prop.get_family(), self.defaultFamily[fontext]))

CPU times: user 1.12 s, sys: 4.64 s, total: 5.76 s
Wall time: 394 ms

	AGYTYP	AGYTYPT	AGY	AGYT	AGYSUB	AGYSUBT
0	1	Cabinet Level Agencies	AF	AF-DEPARTMENT OF THE AIR FORCE	AF**	AF**-INVALID
1	1	Cabinet Level Agencies	AF	AF-DEPARTMENT OF THE AIR FORCE	AF02	AF02-AIR FORCE INSPECTION AGENCY (FO)
2	1	Cabinet Level Agencies	AF	AF-DEPARTMENT OF THE AIR FORCE	AF03	AF03-AIR FORCE OPERATIONAL TEST AND EVALUATION...
3	1	Cabinet Level Agencies	AF	AF-DEPARTMENT OF THE AIR FORCE	AF06	AF06-AIR FORCE AUDIT AGENCY
4	1	Cabinet Level Agencies	AF	AF-DEPARTMENT OF THE AIR FORCE	AF07	AF07-AIR FORCE OFFICE OF SPECIAL INVESTIGATIONS

	QTR	QTRT	EFDATE	EFDATET
0	1	OCT-DEC 2014	201410	OCT 2014
1	1	OCT-DEC 2014	201411	NOV 2014
2	1	OCT-DEC 2014	201412	DEC 2014
3	2	JAN-MAR 2015	201501	JAN 2015
4	2	JAN-MAR 2015	201502	FEB 2015

	GENDER	GENDERT
0	F	Female
1	M	Male
2	Z	Unspecified

	LOCTYP	LOCTYPT	LOC	LOCT
0	1	United States	01	01-ALABAMA
1	1	United States	02	02-ALASKA
2	1	United States	04	04-ARIZONA
3	1	United States	05	05-ARKANSAS
4	1	United States	06	06-CALIFORNIA

	LOSLVL	LOSLVLT
0	A	Less than 1 year
1	B	1 - 2 years
2	C	3 - 4 years
3	D	5 - 9 years
4	E	10 - 14 years

	OCCTYP	OCCTYPT	OCCFAMT	OCC	OCCT
0	1	White Collar	00xx-MISCELLANEOUS OCCUPATIONS	0006	0006-CORRECTIONAL INSTITUTION ADMINISTRATION
1	1	White Collar	00xx-MISCELLANEOUS OCCUPATIONS	0007	0007-CORRECTIONAL OFFICER
2	1	White Collar	00xx-MISCELLANEOUS OCCUPATIONS	0017	0017-EXPLOSIVES SAFETY
3	1	White Collar	00xx-MISCELLANEOUS OCCUPATIONS	0018	0018-SAFETY AND OCCUPATIONAL HEALTH MANAGEMENT
4	1	White Collar	00xx-MISCELLANEOUS OCCUPATIONS	0019	0019-SAFETY TECHNICIAN

	PATCO	PATCOT
0	1	Professional
1	2	Administrative
2	3	Technical
3	4	Clerical
4	5	Other White Collar

	PPTYP	PPTYPT	PPGROUP	PPGROUPT	PAYPLAN	PAYPLANT	PPGRD
0	1	General Schedule and Equivalently Graded (GSEG...	11	Standard GSEG Pay Plans	GL	GL-GS EMPLOYEES IN GRADES 3 THROUGH 10 PAID A ...	GL-03
1	1	General Schedule and Equivalently Graded (GSEG...	11	Standard GSEG Pay Plans	GL	GL-GS EMPLOYEES IN GRADES 3 THROUGH 10 PAID A ...	GL-04
2	1	General Schedule and Equivalently Graded (GSEG...	11	Standard GSEG Pay Plans	GL	GL-GS EMPLOYEES IN GRADES 3 THROUGH 10 PAID A ...	GL-05
3	1	General Schedule and Equivalently Graded (GSEG...	11	Standard GSEG Pay Plans	GL	GL-GS EMPLOYEES IN GRADES 3 THROUGH 10 PAID A ...	GL-06
4	1	General Schedule and Equivalently Graded (GSEG...	11	Standard GSEG Pay Plans	GL	GL-GS EMPLOYEES IN GRADES 3 THROUGH 10 PAID A ...	GL-07

	SALLVL	SALLVLT
0	A	Less than $20,000
1	B	$20,000 - $29,999
2	C	$30,000 - $39,999
3	D	$40,000 - $49,999
4	E	$50,000 - $59,999

	SEP	SEPT
0	SA	Transfer Out - Individual Transfer
1	SB	Transfer Out - Mass Transfer
2	SC	Quit
3	SD	Retirement - Voluntary
4	SE	Retirement - Early Out

	TOATYP	TOATYPT	TOA	TOAT
0	1	Permanent	10	10-Competitive Service - Career
1	1	Permanent	15	15-Competitive Service - Career-Conditional
2	1	Permanent	30	30-Excepted Service - Schedule A
3	1	Permanent	32	32-Excepted Service - Schedule B
4	1	Permanent	34	34-Excepted Service - Schedule C

	WSTYP	WSTYPT	WORKSCH	WORKSCHT
0	1	Full-time	B	B-Full-time Nonseasonal Baylor Plan
1	1	Full-time	F	F-Full-time Nonseasonal
2	1	Full-time	G	G-Full-time Seasonal
3	1	Full-time	H	H-Full-time On-call
4	2	Not Full-time	I	I-Intermittent Nonseasonal

	AGYSUB	SEP	EFDATE	AGELVL	GENDER	GSEGRD	LOSLVL	LOC	OCC	PATCO	PPGRD	SALLVL	TOA	WORKSCH	COUNT	SALARY	LOS
0	AA00	SC	201507	C	M	11	A	11	0905	1	GS-11	F	40	F	1	063722	00.8
1	AA00	SD	201509	K	M	NaN	D	11	0301	2	EX-02	Z	46	F	1	NaN	08.1
2	AA00	SC	201506	D	F	15	C	11	0905	1	GS-15	L	30	F	1	126245	04.8
3	AF**	SA	201503	H	M	11	C	48	2210	2	GS-11	F	10	F	1	066585	04.9
4	AF02	SD	201506	I	M	15	J	35	0301	2	GS-15	O	10	F	1	156737	39.8

	count	mean	std	min	25%	50%	75%	max
COUNT	214282.0	1.000000	0.000000	1.0	1.0	1.0	1.0	1.0
SALARY	214282.0	66479.453855	39471.623281	3913.0	35830.0	54424.0	86910.0	393699.0
LOS	214282.0	11.708865	12.631714	0.0	1.3	6.2	20.4	71.5

	count	mean	std	min	25%	50%	75%	max
SALARY	8008911.0	80067.37279	37918.758366	15120.0	51437.0	74130.0	99957.0	401589.0
LOS	8008911.0	13.06029	10.446755	0.0	4.9	10.0	20.1	71.1

	AGELVL	AGELVLT
0	A	Less than 20
1	B	20-24
2	C	25-29
3	D	30-34
4	E	35-39

	SEP	DATECODE	OCC
217479	NS	201412	7402
217582	NS	201412	7420
217603	NS	201412	1051
217663	NS	201412	1054
218685	NS	201412	2504
218871	NS	201412	8201
218999	NS	201412	4104
219003	NS	201412	4715
219135	NS	201412	0698
220085	NS	201412	0019
220426	NS	201412	3602
221497	NS	201412	2608
221637	NS	201412	3725
224242	NS	201412	6968
225410	NS	201412	0392
226132	NS	201412	3606
228440	NS	201412	2601
231003	NS	201412	3940
231189	NS	201412	5439
246316	NS	201412	1725
246379	NS	201412	5317
246874	NS	201412	5737
247551	NS	201412	1386
254687	NS	201412	0394
259606	NS	201412	4819
264047	NS	201412	2144
266228	NS	201412	1056
268830	NS	201412	5736
270371	NS	201412	0021
271810	NS	201412	3872
271986	NS	201412	4301
273244	NS	201412	3701
273326	NS	201412	6656
273665	NS	201412	8601
275118	NS	201412	3858
275185	NS	201412	4745
277206	NS	201412	4816
279195	NS	201412	1699
284633	NS	201412	5423
289472	NS	201412	1321

	QTR	SEP	OCCT	PPGRD	WORKSCHT
257	4	SC	7401-MISC FOOD PREPARATION AND SERVING	WG-01	Full-time Nonseasonal
627	4	SC	1301-GENERAL PHYSICAL SCIENCE	AD-24	Part-time Nonseasonal
697	4	SJ	0199-SOCIAL SCIENCE STUDENT TRAINEE	GS-02	Intermittent Nonseasonal
749	4	SC	3940-BROADCASTING EQUIPMENT OPERATING	WG-10	Full-time Nonseasonal
2401	4	SJ	0399-ADMINISTRATION AND OFFICE SUPPORT STUDENT...	GS-02	Intermittent Seasonal
3412	2	SC	5003-GARDENING	WG-04	Full-time Seasonal
3471	1	SA	5003-GARDENING	WG-04	Full-time Seasonal
3551	3	SD	5716-ENGINEERING EQUIPMENT OPERATING	WS-14	Full-time Nonseasonal
4937	3	SC	0819-ENVIRONMENTAL ENGINEERING	GS-11	Part-time Job Sharer Nonseasonal
5285	1	SD	5716-ENGINEERING EQUIPMENT OPERATING	WG-08	Intermittent Seasonal
5363	4	SJ	0189-RECREATION AID AND ASSISTANT	GS-03	Intermittent Nonseasonal
5763	1	SD	2005-SUPPLY CLERICAL AND TECHNICIAN	GS-04	Part-time Job Sharer Nonseasonal
6079	3	SC	0180-PSYCHOLOGY	NH-02	Full-time Nonseasonal
6957	3	SD	0810-CIVIL ENGINEERING	DR-03	Full-time Nonseasonal
7015	1	SA	1306-HEALTH PHYSICS	DR-01	Full-time Nonseasonal
7376	4	SC	1699-EQUIPMENT AND FACILITIES MANAGEMENT STUDE...	GS-05	Full-time Nonseasonal
7395	3	SC	0599-FINANCIAL MANAGEMENT STUDENT TRAINEE	DU-01	Full-time Nonseasonal
7464	3	SD	3769-SHOT PEENING MACHINE OPERATING	WS-07	Full-time Nonseasonal
7512	4	SC	0840-NUCLEAR ENGINEERING	DR-03	Full-time Nonseasonal
7675	2	SD	4714-MODEL MAKING	WL-15	Full-time Nonseasonal
7727	4	SC	0189-RECREATION AID AND ASSISTANT	GS-02	Part-time Seasonal
7877	4	SJ	0189-RECREATION AID AND ASSISTANT	GS-02	Part-time Seasonal
8054	4	SC	0665-SPEECH PATHOLOGY AND AUDIOLOGY	DR-03	Full-time Nonseasonal
8160	4	SC	4102-PAINTING	WG-05	Part-time Seasonal
8216	2	SA	5725-CRANE OPERATING	WG-08	Full-time Nonseasonal
8320	4	SD	5401-MISC INDUSTRIAL EQUIPMENT OPERATION	WS-11	Full-time Nonseasonal
8325	1	SC	0189-RECREATION AID AND ASSISTANT	DU-01	Part-time Nonseasonal
8389	4	SD	3705-NON-DESTRUCTIVE TESTING	WS-16	Full-time Nonseasonal
8435	4	SD	1330-ASTRONOMY AND SPACE SCIENCE	DR-04	Full-time Nonseasonal
8449	1	SK	4102-PAINTING	WG-05	Part-time Seasonal
9741	4	SC	0189-RECREATION AID AND ASSISTANT	GS-03	Part-time Seasonal
9890	1	SI	8801-MISCELLANEOUS AIRCRAFT OVERHAUL	WG-08	Part-time Nonseasonal
9903	1	SD	0610-NURSE	DR-01	Full-time Nonseasonal
9916	2	SD	0130-FOREIGN AFFAIRS	DO-02	Full-time Nonseasonal
10139	4	SC	6901-MISC WAREHOUSING AND STOCK HANDLING	WG-06	Part-time Job Sharer Nonseasonal
10656	2	SC	5309-HEATING & BOILER PLANT EQUIPMT MECHANIC	WL-11	Full-time Nonseasonal
11491	4	SC	1008-INTERIOR DESIGN	GG-12	Full-time Nonseasonal
11516	4	SI	0854-COMPUTER ENGINEERING	GG-11	Full-time Nonseasonal
11941	1	SJ	0201-HUMAN RESOURCES MANAGEMENT	GS-06	Full-time Nonseasonal
12265	4	SJ	0335-COMPUTER CLERK AND ASSISTANT	GS-09	Part-time Nonseasonal

	AGYSUB	SEP	DATECODE	AGELVL	GENDER	GSEGRD	LOSLVL	LOC	OCC	PATCO	PPGRD	SALLVL	TOA	WORKSCH	COUNT	SALARY	LOS	AGYTYP	AGYTYPT	AGY	AGYT	AGYSUBT	QTR	AGELVLT	LOSLVLT	LOCTYP	LOCTYPT	LOCT	OCCTYP	OCCTYPT	OCCFAM	OCCFAMT	OCCT	PATCOT	PPTYP	PPTYPT	PPGROUP	PPGROUPT	PAYPLAN	PAYPLANT	SALLVLT	TOATYP	TOATYPT	TOAT	WSTYP	WSTYPT	WORKSCHT	SEPCount_EFDATE_OCC	SEPCount_EFDATE_LOC	IndAvgSalary	SalaryOverUnderIndAvg	LowerLimitAge	YearsToRetirement
8223188	ZU00	NS	201509	D	NaN	NaN	C	11	0301	2	AD-00	G	48	F	NaN	76377.0	4.8	4	Small Independent Agencies (less than 100 empl...	ZU	ZU-DWIGHT D. EISENHOWER MEMORIAL COMMISSION	ZU00-DWIGHT D. EISENHOWER MEMORIAL COMMISSION	4	30-34	3 - 4 years	1	United States	11-DISTRICT OF COLUMBIA	1	White Collar	03	03xx-GENERAL ADMIN, CLERICAL, & OFFICE SVCS	0301-MISCELLANEOUS ADMINISTRATION AND PROGRAM	Administrative	3	Other White Collar Pay Plans	31	Governmentwide or Multi-Agency Plans	AD	AD-ADMINISTRATIVELY DETERMINED RATES, NOT ELSE...	$70,000 - $79,999	2	Non-permanent	48-Excepted Service - Other	1	Full-time	Full-time Nonseasonal	721.0	1391	115840.182250	-39463.182250	30.0	27.0
8223189	ZU00	NS	201509	K	NaN	NaN	D	11	0301	2	AD-00	M	48	F	NaN	139517.0	7.0	4	Small Independent Agencies (less than 100 empl...	ZU	ZU-DWIGHT D. EISENHOWER MEMORIAL COMMISSION	ZU00-DWIGHT D. EISENHOWER MEMORIAL COMMISSION	4	65 or more	5 - 9 years	1	United States	11-DISTRICT OF COLUMBIA	1	White Collar	03	03xx-GENERAL ADMIN, CLERICAL, & OFFICE SVCS	0301-MISCELLANEOUS ADMINISTRATION AND PROGRAM	Administrative	3	Other White Collar Pay Plans	31	Governmentwide or Multi-Agency Plans	AD	AD-ADMINISTRATIVELY DETERMINED RATES, NOT ELSE...	$130,000 - $139,999	2	Non-permanent	48-Excepted Service - Other	1	Full-time	Full-time Nonseasonal	721.0	1391	115840.182250	23676.817750	65.0	-8.0
8223190	ZU00	NS	201509	K	NaN	NaN	D	11	0301	2	AD-00	O	48	F	NaN	158671.0	7.0	4	Small Independent Agencies (less than 100 empl...	ZU	ZU-DWIGHT D. EISENHOWER MEMORIAL COMMISSION	ZU00-DWIGHT D. EISENHOWER MEMORIAL COMMISSION	4	65 or more	5 - 9 years	1	United States	11-DISTRICT OF COLUMBIA	1	White Collar	03	03xx-GENERAL ADMIN, CLERICAL, & OFFICE SVCS	0301-MISCELLANEOUS ADMINISTRATION AND PROGRAM	Administrative	3	Other White Collar Pay Plans	31	Governmentwide or Multi-Agency Plans	AD	AD-ADMINISTRATIVELY DETERMINED RATES, NOT ELSE...	$150,000 - $159,999	2	Non-permanent	48-Excepted Service - Other	1	Full-time	Full-time Nonseasonal	721.0	1391	115840.182250	42830.817750	65.0	-8.0
8223191	ZU00	NS	201509	B	NaN	NaN	B	11	0301	2	AD-00	C	48	F	NaN	36244.0	1.6	4	Small Independent Agencies (less than 100 empl...	ZU	ZU-DWIGHT D. EISENHOWER MEMORIAL COMMISSION	ZU00-DWIGHT D. EISENHOWER MEMORIAL COMMISSION	4	20-24	1 - 2 years	1	United States	11-DISTRICT OF COLUMBIA	1	White Collar	03	03xx-GENERAL ADMIN, CLERICAL, & OFFICE SVCS	0301-MISCELLANEOUS ADMINISTRATION AND PROGRAM	Administrative	3	Other White Collar Pay Plans	31	Governmentwide or Multi-Agency Plans	AD	AD-ADMINISTRATIVELY DETERMINED RATES, NOT ELSE...	$30,000 - $39,999	2	Non-permanent	48-Excepted Service - Other	1	Full-time	Full-time Nonseasonal	721.0	1391	115840.182250	-79596.182250	20.0	37.0
8223192	ZU00	NS	201509	E	NaN	NaN	D	11	0505	2	AD-00	I	48	F	NaN	99288.0	5.0	4	Small Independent Agencies (less than 100 empl...	ZU	ZU-DWIGHT D. EISENHOWER MEMORIAL COMMISSION	ZU00-DWIGHT D. EISENHOWER MEMORIAL COMMISSION	4	35-39	5 - 9 years	1	United States	11-DISTRICT OF COLUMBIA	1	White Collar	05	05xx-ACCOUNTING AND BUDGET	0505-FINANCIAL MANAGEMENT	Administrative	3	Other White Collar Pay Plans	31	Governmentwide or Multi-Agency Plans	AD	AD-ADMINISTRATIVELY DETERMINED RATES, NOT ELSE...	$90,000 - $99,999	2	Non-permanent	48-Excepted Service - Other	1	Full-time	Full-time Nonseasonal	7.0	1391	148382.833333	-49094.833333	35.0	22.0

	QTR	OCC	PPGRD	WORKSCHT	IndAvgSalary
0	1	0006	ES-**	Full-time Nonseasonal	161827.273973
1	1	0006	GL-09	Full-time Nonseasonal	63970.126984
2	1	0006	GS-09	Full-time Nonseasonal	56876.500000
3	1	0006	GS-11	Full-time Nonseasonal	72865.783673
4	1	0006	GS-12	Full-time Nonseasonal	85742.663717

	AGYSUB	SEP	DATECODE	AGELVL	GENDER	GSEGRD	LOSLVL	LOC	OCC	PATCO	PPGRD	SALLVL	TOA	WORKSCH	COUNT	SALARY	LOS	AGYTYP	AGYTYPT	AGY	AGYT	AGYSUBT	QTR	AGELVLT	LOSLVLT	LOCTYP	LOCTYPT	LOCT	OCCTYP	OCCTYPT	OCCFAM	OCCFAMT	OCCT	PATCOT	PPTYP	PPTYPT	PPGROUP	PPGROUPT	PAYPLAN	PAYPLANT	SALLVLT	TOATYP	TOATYPT	TOAT	WSTYP	WSTYPT	WORKSCHT	SEPCount_EFDATE_OCC	SEPCount_EFDATE_LOC	IndAvgSalary	SalaryOverUnderIndAvg	LowerLimitAge	YearsToRetirement	BLS_FEDERAL_OtherSep_Rate	BLS_FEDERAL_Quits_Rate	BLS_FEDERAL_TotalSep_Level	BLS_FEDERAL_JobOpenings_Rate	BLS_FEDERAL_OtherSep_Level	BLS_FEDERAL_Quits_Level	BLS_FEDERAL_JobOpenings_Level	BLS_FEDERAL_Layoffs_Rate	BLS_FEDERAL_Layoffs_Level	BLS_FEDERAL_TotalSep_Rate
8170902	ZU00	NS	201509	D	NaN	NaN	C	11	0301	2	AD-00	G	48	F	NaN	76377.0	4.8	4	Small Independent Agencies (less than 100 empl...	ZU	ZU-DWIGHT D. EISENHOWER MEMORIAL COMMISSION	ZU00-DWIGHT D. EISENHOWER MEMORIAL COMMISSION	4	30-34	3 - 4 years	1	United States	11-DISTRICT OF COLUMBIA	1	White Collar	03	03xx-GENERAL ADMIN, CLERICAL, & OFFICE SVCS	0301-MISCELLANEOUS ADMINISTRATION AND PROGRAM	Administrative	3	Other White Collar Pay Plans	31	Governmentwide or Multi-Agency Plans	AD	AD-ADMINISTRATIVELY DETERMINED RATES, NOT ELSE...	$70,000 - $79,999	2	Non-permanent	48-Excepted Service - Other	1	Full-time	Full-time Nonseasonal	721.0	1391	115840.182250	-39463.182250	30.0	27.0	0.4	0.5	38	1.9	12	14	55	0.4	12	1.4
8170903	ZU00	NS	201509	K	NaN	NaN	D	11	0301	2	AD-00	M	48	F	NaN	139517.0	7.0	4	Small Independent Agencies (less than 100 empl...	ZU	ZU-DWIGHT D. EISENHOWER MEMORIAL COMMISSION	ZU00-DWIGHT D. EISENHOWER MEMORIAL COMMISSION	4	65 or more	5 - 9 years	1	United States	11-DISTRICT OF COLUMBIA	1	White Collar	03	03xx-GENERAL ADMIN, CLERICAL, & OFFICE SVCS	0301-MISCELLANEOUS ADMINISTRATION AND PROGRAM	Administrative	3	Other White Collar Pay Plans	31	Governmentwide or Multi-Agency Plans	AD	AD-ADMINISTRATIVELY DETERMINED RATES, NOT ELSE...	$130,000 - $139,999	2	Non-permanent	48-Excepted Service - Other	1	Full-time	Full-time Nonseasonal	721.0	1391	115840.182250	23676.817750	65.0	-8.0	0.4	0.5	38	1.9	12	14	55	0.4	12	1.4
8170904	ZU00	NS	201509	K	NaN	NaN	D	11	0301	2	AD-00	O	48	F	NaN	158671.0	7.0	4	Small Independent Agencies (less than 100 empl...	ZU	ZU-DWIGHT D. EISENHOWER MEMORIAL COMMISSION	ZU00-DWIGHT D. EISENHOWER MEMORIAL COMMISSION	4	65 or more	5 - 9 years	1	United States	11-DISTRICT OF COLUMBIA	1	White Collar	03	03xx-GENERAL ADMIN, CLERICAL, & OFFICE SVCS	0301-MISCELLANEOUS ADMINISTRATION AND PROGRAM	Administrative	3	Other White Collar Pay Plans	31	Governmentwide or Multi-Agency Plans	AD	AD-ADMINISTRATIVELY DETERMINED RATES, NOT ELSE...	$150,000 - $159,999	2	Non-permanent	48-Excepted Service - Other	1	Full-time	Full-time Nonseasonal	721.0	1391	115840.182250	42830.817750	65.0	-8.0	0.4	0.5	38	1.9	12	14	55	0.4	12	1.4
8170905	ZU00	NS	201509	B	NaN	NaN	B	11	0301	2	AD-00	C	48	F	NaN	36244.0	1.6	4	Small Independent Agencies (less than 100 empl...	ZU	ZU-DWIGHT D. EISENHOWER MEMORIAL COMMISSION	ZU00-DWIGHT D. EISENHOWER MEMORIAL COMMISSION	4	20-24	1 - 2 years	1	United States	11-DISTRICT OF COLUMBIA	1	White Collar	03	03xx-GENERAL ADMIN, CLERICAL, & OFFICE SVCS	0301-MISCELLANEOUS ADMINISTRATION AND PROGRAM	Administrative	3	Other White Collar Pay Plans	31	Governmentwide or Multi-Agency Plans	AD	AD-ADMINISTRATIVELY DETERMINED RATES, NOT ELSE...	$30,000 - $39,999	2	Non-permanent	48-Excepted Service - Other	1	Full-time	Full-time Nonseasonal	721.0	1391	115840.182250	-79596.182250	20.0	37.0	0.4	0.5	38	1.9	12	14	55	0.4	12	1.4
8170906	ZU00	NS	201509	E	NaN	NaN	D	11	0505	2	AD-00	I	48	F	NaN	99288.0	5.0	4	Small Independent Agencies (less than 100 empl...	ZU	ZU-DWIGHT D. EISENHOWER MEMORIAL COMMISSION	ZU00-DWIGHT D. EISENHOWER MEMORIAL COMMISSION	4	35-39	5 - 9 years	1	United States	11-DISTRICT OF COLUMBIA	1	White Collar	05	05xx-ACCOUNTING AND BUDGET	0505-FINANCIAL MANAGEMENT	Administrative	3	Other White Collar Pay Plans	31	Governmentwide or Multi-Agency Plans	AD	AD-ADMINISTRATIVELY DETERMINED RATES, NOT ELSE...	$90,000 - $99,999	2	Non-permanent	48-Excepted Service - Other	1	Full-time	Full-time Nonseasonal	7.0	1391	148382.833333	-49094.833333	35.0	22.0	0.4	0.5	38	1.9	12	14	55	0.4	12	1.4

	SEP	StratCount
0	NS	7957918
1	SA	26945
2	SB	333
3	SC	66248
4	SD	56820
5	SE	1260
6	SF	4100
7	SG	1467
8	SH	400
9	SI	9728
10	SJ	42754
11	SK	2892
12	SL	42

	SEP_SA	SEP_SC	SEP_SD
0	0	1	0
1	0	1	0
2	1	0	0
3	0	0	1
4	0	1	0

	SALARY	LOS	SEPCount_EFDATE_OCC	SEPCount_EFDATE_LOC	IndAvgSalary	SalaryOverUnderIndAvg	LowerLimitAge	YearsToRetirement	SEP	SEP_SA	SEP_SC	SEP_SD
0	63722.0	0.8	205.0	1319	64540.593830	-818.593830	25.0	32.0	SC	0	1	0
1	126245.0	4.8	207.0	1132	149864.298504	-23619.298504	30.0	27.0	SC	0	1	0
2	66585.0	4.9	439.0	1087	71530.963755	-4945.963755	50.0	7.0	SA	1	0	0
3	156737.0	39.8	670.0	265	146735.220304	10001.779696	55.0	2.0	SD	0	0	1
4	92973.0	1.0	721.0	1853	101641.124025	-8668.124025	50.0	7.0	SC	0	1	0

	SEP	DATECODE	AGELVL	GSEGRD	LOC	PATCO	TOA	WORKSCH	SALARY	LOS	AGELVLT	LOCTYP	LOCTYPT	LOCT	OCCTYP	OCCTYPT	OCCFAM	OCCFAMT	PATCOT	PPTYP	PPTYPT	PPGROUP	PPGROUPT	TOATYP	TOATYPT	TOAT	WSTYP	WSTYPT	WORKSCHT	SEPCount_EFDATE_OCC	SEPCount_EFDATE_LOC	IndAvgSalary	SalaryOverUnderIndAvg	LowerLimitAge	YearsToRetirement	BLS_FEDERAL_OtherSep_Rate	BLS_FEDERAL_Quits_Rate	BLS_FEDERAL_TotalSep_Level	BLS_FEDERAL_JobOpenings_Rate	BLS_FEDERAL_OtherSep_Level	BLS_FEDERAL_Quits_Level	BLS_FEDERAL_JobOpenings_Level	BLS_FEDERAL_Layoffs_Rate	BLS_FEDERAL_Layoffs_Level	BLS_FEDERAL_TotalSep_Rate	SALARYLog	LOSSqrt	SEPCount_EFDATE_OCCLog	SEPCount_EFDATE_LOCLog	IndAvgSalaryLog
0	SC	201507	C	11.0	11	1	40	F	63722.0	0.8	25-29	1	United States	11-DISTRICT OF COLUMBIA	1	White Collar	09	09xx-LEGAL AND KINDRED	Professional	1	General Schedule and Equivalently Graded (GSEG...	11	Standard GSEG Pay Plans	2	Non-permanent	40-Excepted Service - Schedule A	1	Full-time	Full-time Nonseasonal	205.0	1319	64540.593830	-818.593830	25.0	32.0	0.4	0.5	34	2.6	11	13	74	0.4	10	1.2	11.062285	0.894427	5.323010	7.184629	11.075050
1	SC	201506	D	15.0	11	1	30	F	126245.0	4.8	30-34	1	United States	11-DISTRICT OF COLUMBIA	1	White Collar	09	09xx-LEGAL AND KINDRED	Professional	1	General Schedule and Equivalently Graded (GSEG...	11	Standard GSEG Pay Plans	1	Permanent	30-Excepted Service - Schedule A	1	Full-time	Full-time Nonseasonal	207.0	1132	149864.298504	-23619.298504	30.0	27.0	0.4	0.5	34	2.3	12	13	65	0.4	10	1.2	11.745980	2.190890	5.332719	7.031741	11.917485
8	SD	201509	I	14.0	06	1	10	F	135500.0	14.3	55-59	1	United States	06-CALIFORNIA	1	White Collar	08	08xx-ENGINEERING AND ARCHITECTURE	Professional	1	General Schedule and Equivalently Graded (GSEG...	11	Standard GSEG Pay Plans	1	Permanent	10-Competitive Service - Career	1	Full-time	Full-time Nonseasonal	122.0	1853	125803.916312	9696.083688	55.0	2.0	0.4	0.5	38	1.9	12	14	55	0.4	12	1.4	11.816727	3.781534	4.804021	7.524561	11.742480
11	SD	201503	J	14.0	08	1	10	F	128223.0	20.6	60-64	1	United States	08-COLORADO	1	White Collar	08	08xx-ENGINEERING AND ARCHITECTURE	Professional	1	General Schedule and Equivalently Graded (GSEG...	11	Standard GSEG Pay Plans	1	Permanent	10-Competitive Service - Career	1	Full-time	Full-time Nonseasonal	92.0	329	126328.546349	1894.453651	60.0	-3.0	0.3	0.4	31	3.0	9	10	86	0.5	12	1.1	11.761526	4.538722	4.521789	5.796058	11.746641
14	SA	201508	H	13.0	06	1	10	F	111566.0	24.3	50-54	1	United States	06-CALIFORNIA	1	White Collar	08	08xx-ENGINEERING AND ARCHITECTURE	Professional	1	General Schedule and Equivalently Graded (GSEG...	11	Standard GSEG Pay Plans	1	Permanent	10-Competitive Service - Career	1	Full-time	Full-time Nonseasonal	110.0	1606	105047.296509	6518.703491	50.0	7.0	0.5	0.6	41	2.3	14	17	67	0.3	10	1.5	11.622372	4.929503	4.700480	7.381502	11.562166

	SEP	StratCount
0	NS	4101470
1	SA	17983
2	SB	257
3	SC	22021
4	SD	35304
5	SE	876
6	SF	1988
7	SG	788
8	SH	79
9	SI	2727
10	SJ	5926
11	SK	1575
12	SL	18

	SEP	StratCount	StratCountSample
0	NS	1259283	7500.0
1	SA	5463	5463.0
2	SB	85	85.0
3	SC	7423	7423.0
4	SD	8881	7500.0
5	SE	167	167.0
6	SF	338	338.0
7	SG	90	90.0
8	SH	15	15.0
9	SI	631	631.0
10	SJ	645	645.0
11	SK	355	355.0
12	SL	7	7.0

	SEP	StratCount	StratCountSample
0	NS	2087084	7500.0
1	SA	9252	7500.0
2	SB	145	145.0
3	SC	9156	7500.0
4	SD	19402	7500.0
5	SE	555	555.0
6	SF	995	995.0
7	SG	414	414.0
8	SH	39	39.0
9	SI	1196	1196.0
10	SJ	2771	2771.0
11	SK	820	820.0
12	SL	11	11.0

	SEP	DATECODE	AGELVL	GSEGRD	LOC	PATCO	TOA	WORKSCH	SALARY	LOS	AGELVLT	LOCTYP	LOCTYPT	LOCT	OCCTYP	OCCTYPT	OCCFAM	OCCFAMT	PATCOT	PPTYP	PPTYPT	PPGROUP	PPGROUPT	TOATYP	TOATYPT	TOAT	WSTYP	WSTYPT	WORKSCHT	SEPCount_EFDATE_OCC	SEPCount_EFDATE_LOC	IndAvgSalary	SalaryOverUnderIndAvg	LowerLimitAge	YearsToRetirement	BLS_FEDERAL_OtherSep_Rate	BLS_FEDERAL_Quits_Rate	BLS_FEDERAL_TotalSep_Level	BLS_FEDERAL_JobOpenings_Rate	BLS_FEDERAL_OtherSep_Level	BLS_FEDERAL_Quits_Level	BLS_FEDERAL_JobOpenings_Level	BLS_FEDERAL_Layoffs_Rate	BLS_FEDERAL_Layoffs_Level	BLS_FEDERAL_TotalSep_Rate	SALARYLog	LOSSqrt	SEPCount_EFDATE_OCCLog	SEPCount_EFDATE_LOCLog	IndAvgSalaryLog
0	NS	201412	B	11.0	51	1	15	F	77658.0	1.5	20-24	1	United States	51-VIRGINIA	1	White Collar	12	12xx-COPYRIGHT, PATENT, AND TRADE-MARK	Professional	1	General Schedule and Equivalently Graded (GSEG...	11	Standard GSEG Pay Plans	1	Permanent	15-Competitive Service - Career-Conditional	1	Full-time	Full-time Nonseasonal	26.0	1133	78919.462629	-1261.462629	20.0	37.0	0.5	0.4	30	2.2	12	10	62	0.3	7	1.1	11.260070	1.224745	3.258097	7.032624	11.276183
1	NS	201412	B	9.0	11	1	15	F	52146.0	2.9	20-24	1	United States	11-DISTRICT OF COLUMBIA	1	White Collar	11	11xx-BUSINESS AND INDUSTRY	Professional	1	General Schedule and Equivalently Graded (GSEG...	11	Standard GSEG Pay Plans	1	Permanent	15-Competitive Service - Career-Conditional	1	Full-time	Full-time Nonseasonal	274.0	1260	52460.545881	-314.545881	20.0	37.0	0.5	0.4	30	2.2	12	10	62	0.3	7	1.1	10.861803	1.702939	5.613128	7.138867	10.867817
2	NS	201412	B	9.0	53	1	15	F	57368.0	0.5	20-24	1	United States	53-WASHINGTON	1	White Collar	08	08xx-ENGINEERING AND ARCHITECTURE	Professional	1	General Schedule and Equivalently Graded (GSEG...	11	Standard GSEG Pay Plans	1	Permanent	15-Competitive Service - Career-Conditional	1	Full-time	Full-time Nonseasonal	7.0	555	58563.540541	-1195.540541	20.0	37.0	0.5	0.4	30	2.2	12	10	62	0.3	7	1.1	10.957242	0.707107	1.945910	6.318968	10.977868
3	NS	201412	B	7.0	25	1	15	F	42830.0	1.5	20-24	1	United States	25-MASSACHUSETTS	1	White Collar	05	05xx-ACCOUNTING AND BUDGET	Professional	1	General Schedule and Equivalently Graded (GSEG...	11	Standard GSEG Pay Plans	1	Permanent	15-Competitive Service - Career-Conditional	1	Full-time	Full-time Nonseasonal	82.0	304	43211.010135	-381.010135	20.0	37.0	0.5	0.4	30	2.2	12	10	62	0.3	7	1.1	10.664994	1.224745	4.406719	5.717028	10.673851
4	NS	201412	B	7.0	53	1	15	F	44615.0	2.5	20-24	1	United States	53-WASHINGTON	1	White Collar	08	08xx-ENGINEERING AND ARCHITECTURE	Professional	1	General Schedule and Equivalently Graded (GSEG...	11	Standard GSEG Pay Plans	1	Permanent	15-Competitive Service - Career-Conditional	1	Full-time	Full-time Nonseasonal	50.0	555	49854.574586	-5239.574586	20.0	37.0	0.5	0.4	30	2.2	12	10	62	0.3	7	1.1	10.705825	1.581139	3.912023	6.318968	10.816866

	SEP_SA	SEP_SC	SEP_SD
0	0	1	0
1	0	1	0
2	1	0	0
3	0	0	1
4	0	1	0

	SEP	StratCount
0	NS	7501
1	SA	5463
2	SB	85
3	SC	7423
4	SD	7507
5	SE	167
6	SF	338
7	SG	90
8	SH	15
9	SI	631
10	SJ	645
11	SK	355
12	SL	7

	SEP	DATECODE	AGELVL	GSEGRD	LOC	PATCO	TOA	WORKSCH	SALARY	LOS	AGELVLT	LOCTYP	LOCTYPT	LOCT	OCCTYP	OCCTYPT	OCCFAM	OCCFAMT	PATCOT	PPTYP	PPTYPT	PPGROUP	PPGROUPT	TOATYP	TOATYPT	TOAT	WSTYP	WSTYPT	WORKSCHT	SEPCount_EFDATE_OCC	SEPCount_EFDATE_LOC	IndAvgSalary	SalaryOverUnderIndAvg	LowerLimitAge	YearsToRetirement	BLS_FEDERAL_OtherSep_Rate	BLS_FEDERAL_Quits_Rate	BLS_FEDERAL_TotalSep_Level	BLS_FEDERAL_JobOpenings_Rate	BLS_FEDERAL_OtherSep_Level	BLS_FEDERAL_Quits_Level	BLS_FEDERAL_JobOpenings_Level	BLS_FEDERAL_Layoffs_Rate	BLS_FEDERAL_Layoffs_Level	BLS_FEDERAL_TotalSep_Rate	SALARYLog	LOSSqrt	SEPCount_EFDATE_OCCLog	SEPCount_EFDATE_LOCLog	IndAvgSalaryLog
0	NS	201412	B	7.0	11	2	38	F	42631.0	0.1	20-24	1	United States	11-DISTRICT OF COLUMBIA	1	White Collar	03	03xx-GENERAL ADMIN, CLERICAL, & OFFICE SVCS	Administrative	1	General Schedule and Equivalently Graded (GSEG...	11	Standard GSEG Pay Plans	1	Permanent	38-Excepted Service - Other	1	Full-time	Full-time Nonseasonal	774.0	1260	44891.572603	-2260.572603	20.0	37.0	0.5	0.4	30	2.2	12	10	62	0.3	7	1.1	10.660337	0.316228	6.651572	7.138867	10.712005
1	NS	201412	B	7.0	51	2	15	F	42631.0	5.0	20-24	1	United States	51-VIRGINIA	1	White Collar	11	11xx-BUSINESS AND INDUSTRY	Administrative	1	General Schedule and Equivalently Graded (GSEG...	11	Standard GSEG Pay Plans	1	Permanent	15-Competitive Service - Career-Conditional	1	Full-time	Full-time Nonseasonal	14.0	1133	45621.916667	-2990.916667	20.0	37.0	0.5	0.4	30	2.2	12	10	62	0.3	7	1.1	10.660337	2.236068	2.639057	7.032624	10.728144
2	NS	201412	B	9.0	51	2	10	F	48893.0	3.4	20-24	1	United States	51-VIRGINIA	1	White Collar	03	03xx-GENERAL ADMIN, CLERICAL, & OFFICE SVCS	Administrative	1	General Schedule and Equivalently Graded (GSEG...	11	Standard GSEG Pay Plans	1	Permanent	10-Competitive Service - Career	1	Full-time	Full-time Nonseasonal	774.0	1133	57175.047502	-8282.047502	20.0	37.0	0.5	0.4	30	2.2	12	10	62	0.3	7	1.1	10.797390	1.843909	6.651572	7.032624	10.953873
3	NS	201412	B	9.0	26	2	38	F	53828.0	1.7	20-24	1	United States	26-MICHIGAN	1	White Collar	21	21xx-TRANSPORTATION	Administrative	1	General Schedule and Equivalently Graded (GSEG...	11	Standard GSEG Pay Plans	1	Permanent	38-Excepted Service - Other	1	Full-time	Full-time Nonseasonal	80.0	214	56335.117117	-2507.117117	20.0	37.0	0.5	0.4	30	2.2	12	10	62	0.3	7	1.1	10.893549	1.303840	4.382027	5.365976	10.939073
4	NS	201412	B	7.0	34	2	48	F	41797.0	0.8	20-24	1	United States	34-NEW JERSEY	1	White Collar	00	00xx-MISCELLANEOUS OCCUPATIONS	Administrative	1	General Schedule and Equivalently Graded (GSEG...	11	Standard GSEG Pay Plans	2	Non-permanent	48-Excepted Service - Other	1	Full-time	Full-time Nonseasonal	113.0	233	44956.060811	-3159.060811	20.0	37.0	0.5	0.4	30	2.2	12	10	62	0.3	7	1.1	10.640580	0.894427	4.727388	5.451038	10.713441

	SEP	StratCount
0	NS	7498
1	SA	7501
2	SB	145
3	SC	7507
4	SD	7504
5	SE	555
6	SF	995
7	SG	414
8	SH	39
9	SI	1196
10	SJ	2771
11	SK	820
12	SL	11

	count	mean	std	min	25%	50%	75%	max
GSEGRD	30227.0	12.142588	1.757041	7.000000	11.000000	12.000000	13.000000	15.000000
SALARY	30227.0	95587.756278	31018.478638	39179.000000	73143.000000	90711.000000	113673.000000	364782.000000
LOS	30227.0	14.333047	12.159974	0.000000	4.600000	9.700000	24.800000	71.500000
SEPCount_EFDATE_OCC	30227.0	146.208820	161.039507	1.000000	32.000000	82.000000	207.000000	708.000000
SEPCount_EFDATE_LOC	30227.0	737.081086	496.100839	18.000000	310.000000	596.000000	1123.000000	2791.000000
IndAvgSalary	30227.0	94981.435573	29622.413408	39179.000000	70499.260612	86256.027972	106119.623016	220806.750000
SalaryOverUnderIndAvg	30227.0	606.320705	9947.104620	-119176.750000	-5362.868080	227.033473	6789.125828	147896.670738
LowerLimitAge	30227.0	46.216462	12.931906	20.000000	35.000000	50.000000	60.000000	65.000000
YearsToRetirement	30227.0	10.783538	12.931906	-8.000000	-3.000000	7.000000	22.000000	37.000000
BLS_FEDERAL_OtherSep_Rate	30227.0	0.429315	0.084823	0.300000	0.400000	0.400000	0.500000	0.600000
BLS_FEDERAL_Quits_Rate	30227.0	0.455067	0.068862	0.300000	0.400000	0.500000	0.500000	0.600000
BLS_FEDERAL_TotalSep_Level	30227.0	36.292950	8.526718	26.000000	31.000000	34.000000	38.000000	60.000000
BLS_FEDERAL_JobOpenings_Rate	30227.0	2.450402	0.390278	1.900000	2.200000	2.300000	2.800000	3.200000
BLS_FEDERAL_OtherSep_Level	30227.0	11.951368	2.383434	8.000000	10.000000	12.000000	12.000000	17.000000
BLS_FEDERAL_Quits_Level	30227.0	12.004565	2.058074	9.000000	10.000000	13.000000	13.000000	17.000000
BLS_FEDERAL_JobOpenings_Level	30227.0	69.714196	11.375094	55.000000	62.000000	67.000000	80.000000	91.000000
BLS_FEDERAL_Layoffs_Rate	30227.0	0.461144	0.216841	0.300000	0.400000	0.400000	0.500000	1.100000
BLS_FEDERAL_Layoffs_Level	30227.0	12.412942	5.974603	7.000000	10.000000	12.000000	12.000000	30.000000
BLS_FEDERAL_TotalSep_Rate	30227.0	1.314652	0.316763	1.000000	1.100000	1.200000	1.400000	2.200000
SALARYLog	30227.0	11.417742	0.315717	10.575896	11.200172	11.415434	11.641081	12.807055
LOSSqrt	30227.0	3.392935	1.679622	0.000000	2.144761	3.114482	4.979960	8.455767
SEPCount_EFDATE_OCCLog	30227.0	4.297617	1.332665	0.000000	3.465736	4.406719	5.332719	6.562444
SEPCount_EFDATE_LOCLog	30227.0	6.319419	0.823629	2.890372	5.736572	6.390241	7.023759	7.934155
IndAvgSalaryLog	30227.0	11.416091	0.298912	10.575896	11.163358	11.365075	11.572322	12.305043

	SEP_NS	SEP_SA	SEP_SB	SEP_SC	SEP_SD	SEP_SE	SEP_SF	SEP_SG	SEP_SH	SEP_SI	SEP_SJ	SEP_SK	SEP_SL
0	1	0	0	0	0	0	0	0	0	0	0	0	0
1	1	0	0	0	0	0	0	0	0	0	0	0	0
2	1	0	0	0	0	0	0	0	0	0	0	0	0
3	1	0	0	0	0	0	0	0	0	0	0	0	0
4	1	0	0	0	0	0	0	0	0	0	0	0	0

	DATECODE	AGELVL	StratCount	SEP	TotalCount	p	StratCountSample	StratSampleSize
0	201410	B	2	NS	21076	0.000095	7500.0	1
1	201410	C	57	NS	21076	0.002704	7500.0	20
2	201410	D	85	NS	21076	0.004033	7500.0	30
3	201410	E	73	NS	21076	0.003464	7500.0	26
4	201410	F	74	NS	21076	0.003511	7500.0	26

	DATECODE	AGELVL	StratCount	SEP	TotalCount	p	StratCountSample	StratSampleSize
0	201410	B	10	SC	7423	0.001347	7423.0	10
1	201410	C	92	SC	7423	0.012394	7423.0	92
2	201410	D	154	SC	7423	0.020746	7423.0	154
3	201410	E	118	SC	7423	0.015897	7423.0	118
4	201410	F	80	SC	7423	0.010777	7423.0	80

	DATECODE	AGELVL	StratCount	SEP	TotalCount	p	StratCountSample	StratSampleSize
0	201410	B	3	NS	24237	0.000124	7500.0	1
1	201410	C	57	NS	24237	0.002352	7500.0	18
2	201410	D	133	NS	24237	0.005487	7500.0	41
3	201410	E	98	NS	24237	0.004043	7500.0	30
4	201410	F	114	NS	24237	0.004704	7500.0	35

	DATECODE	AGELVL	StratCount	SEP	TotalCount	p	StratCountSample	StratSampleSize
0	201410	B	16	SC	7507	0.002131	7423.0	16
1	201410	C	77	SC	7507	0.010257	7423.0	76
2	201410	D	142	SC	7507	0.018916	7423.0	140
3	201410	E	110	SC	7507	0.014653	7423.0	109
4	201410	F	66	SC	7507	0.008792	7423.0	65